Skip to main content

Full text of "The Biological bulletin"

See other formats 


THE 


BIOLOGICAL  BULLETIN 


PUBLISHED  BY 

THE  MARINE  BIOLOGICAL  LABORATORY 


Editorial  Board 

JOHN  M.  ANDERSON,  Cornell  University  MEREDITH  L.  JONES,  Smithsonian  Institution 

JOHN  B.  BUCK,  National  Institutes  of  Health  GEORGE  O.  MACKIE,  University  of  Victoria 

RALPH  I.  SMITH,  University  of  California, 

DONALD  P.  COSTELLO,  Woods  Hole,  Berkeley 

Massachusetts 

F.  JOHN  VERNBERG,  University  of 
JOHN  D.  COSTLOW,  Duke  University  South  Carolina 

PHILIP  B.  DUNHAM,  Syracuse  University  CARROLL  M.  WILLIAMS,  Harvard  University 

CATHERINE  HENLEY,  National  Institutes  of  Health      EDWARD  O.  WILSON,  Harvard  University 

W.  D.  RUSSELL-HUNTER,  Syracuse  University 
Managing  Editor 


VOLUME  154 

FEBRUARY  TO  JUNE,  1978 


Printed  and  Issued  by 

LANCASTER  PRESS,  Inc. 

PRINCE  £  LEMON  STS. 

LANCASTER,  PA. 


THE  BIOLOGICAL  BULLETIN  is  issued  six  times  a  year  at  the 
Lancaster  Press,  Inc.,  Prince  and  Lemon  Streets,  Lancaster,  Penn- 
sylvania. 

Subscriptions  and  similar  matter  should  be  addressed  to  The 
Biological  Bulletin,  Marine  Biological  Laboratory,  Woods  Hole, 
Massachusetts.  Agent  for  Great  Britain :  Wheldon  and  Wesley, 
Limited,  2,  3  and  4  Arthur  Street,  New  Oxford  Street,  London, 
W.  C.  2.  Single  numbers,  $8.00.  Subscription  per  volume  (three 
issues),  $22.00. 

Communications  relative  to  manuscripts  should  be  sent  to  Dr. 
W.  D.  Russell-Hunter,  Marine  Biological  Laboratory,  Woods 
Hole,  Massachusetts  02543  between  June  1  and  September  1, 
and  to  Dr.  W.  D.  Russell-Hunter,  P.O.  Box  103,  University 
Station,  Syracuse,  New  York  13210,  during  the  remainder  of 
the  year. 


Second-class-postage  paid  at  Lancaster,  Pa. 


LANCASTER  PRESS,  INC.,  LANCASTER,  PA. 


CONTENTS 

No.  1,  FEBRUARY,  1978 

ANDERSON,  JOHN  MAXWELL 

Studies  on  functional  morphology  in  the  digestive  system  of  Oreaster 
reticulatus  (L.)  (Asteroidea) 1 

ARMSTRONG,  DAVID  A.,   DEBBIE  CHIPPENDALE,  ALLEN  W.  KNIGHT  AND 
JOHN  E.  COLT 

Interaction  of  ionized  and  un-ionized  ammonia  on  short-term  survival 
and  growth  of  prawn  larvae,  Macrobrachium  rosenbergii 15 

BARKER,  M.  F. 

Descriptions  of  the  larvae  of  Stichaster  austral  is  (Verrill)  and  Cosci- 
nasterias  calamaria  (Gray)  (Echinodermata:  Asteroidea)  from  New 
Zealand,  obtained  from  laboratory  culture 32 

CONKLIN,  D.   E.  AND  L.  PROVASOLI 

Diphasic  particulate  media  for  the  culture  of  filter  feeders 47 

GOVIND,  C.  K.  AND  FRED  LANG 

Development  of  the  dimorphic  claw  closer  muscles  of  the  lobster 
Homarus  americanus.  III.  Transformation  to  dimorphic  muscles  in 
juveniles 55 

GREEN,  JEFFREY  D. 

The  annual  reproductive  cycle  of  an  apodous  holothurian,  Lepto- 
synapta  tennis:  a  bimodal  breeding  season 68 

HENDLER,  GORDON 

Development  of  Amphioplus  abditus  (Yerrill)  (Echinodermata: 
Ophiuroidea).  II.  Description  and  discussion  of  ophiuroid  skeletal 
ontogeny  and  homologies 79 

HOVE,  H.  A.  TEN  AND  J.  C.  A.  WEERDENBURG 

A  generic  revision  of  the  brackish-water  serpulid  Ficopomatiis 
Southern  1921  (Polychaeta:  Serpulinae),  including  Mercierella  Fauvel 
1923,  Sphaeropomatus  Treadwell  1934,  Mercierellopsis  Rioja  1945  and 
Neopomatus  Pillai  1960 96 

KURIS,  ARMAND  M. 

Life  cycle,  distribution  and  abundance  of  Carcinonemertes  epialti,  a 
nemertean  egg  predator  of  the  shore  crab  Hemigrapsus  oregonensis, 
in  relation  to  host  size,  reproduction,  and  molt  cycle 121 

MICKEL,  T.  J.  AND  J.  J.  CHILDRESS 

The  effect  of  pH  on  oxygen  consumption  and  activity  in  the  bathy- 
pelagic  mysid  Gnathophausia  ingens 138 

PEZALLA,  PAUL  D.,  ROBERT  M.  DORES  AND  WILLIAM  S.  HERMAN 

Separation  and  partial  purification  of  central  nervous  system  peptides 
from  Limulus  polyphemus  with  hyperglycemic  and  chromatophoro- 
tropic  activity  in  crustaceans 148 

RIVEST,  BRIAN  R. 

Development  of  the  eolid  nudibranch  Cuthona  nana  (Alder  and 
Hancock,  1842),  and  its  relationship  with  a  hydroid  and  hermit  crab  157 


in 


iv  CONTENTS 

No.  2,  APRIL,  1978 

BRADLEY,  BRIAN  P. 

Increase  in  range  of  temperature  tolerance  by  acclimation  in  the 
copepod  Eurytemora  affinis 177 

DAME,  R.  F.  AND  F.  J.  VERNBERG 

The  influence  of  constant  and  cyclic  acclimation  temperatures  on  the 
metabolic  rates  of  Panopeus  herbstii  and  Uca  pugilator 188 

DEL  PINO,  EUGENIA  M.,  AND  A.  A.  HUMPHRIES,  JR. 

Multiple  nuclei  during  early  oogenesis  in  Flectonotns  pygmaeiis  and 
other  marsupial  frogs 198 

FISHER,  FRANK  M.,  JR.  AND  JOHN  A.  OAKS 

Evidence  for  a  nonintestinal  nutritional  mechanism  in  the  rhyn- 
chocoelan,  Linens  ruber 213 

FUZESSERY,  ZOLTAN  M.,  WlLLIAM  E.  S.  CARR,  AND  BARRY  W.  ACHE 

Antennular  chemosensitivity  in  the  spiny  lobster,  Panulirns  argus: 
studies  of  taurine  sensitive  receptors 226 

GOY,  JOSEPH  W.  AND  ANTHONY  J.  PROVENZANO,  JR. 

Larval  development  of  the  rare  burrowing  mud  shrimp  Naushonia 
crangonoides  Kingsley  (Decapoda:  Thalassinidea;  Laomediidae) 241 

HINES,  ANSON  H. 

Reproduction  in  three  species  of  intertidal  barnacles  from  central 
California 262 

PECHENIK,  JAN  A. 

Adaptations  to  intertidal  development:  studies  on  Nassarius  obsoletus     282 

PRUSCH,  ROBERT  D.  AND  CAROL  HALL 

Diffusional  water  permeability  in  selected  marine  bivalves 292 

ROBERTSON,  DOUGLAS  R. 

The  light-dark  cycle  and  a  nonlinear  analysis  of  lunar  perturbations 
and  barometric  pressure  associated  with  the  annual  locomotor  activity 
of  the  frog,  Rana  pipiens 302 

SHIRLEY,  THOMAS  C.,  GUY  J.  DENOUX,  AND  WILLIAM  B.  STICKLE 

Seasonal  respiration  in  the  marsh  periwinkle,  Littorina  irrorata 322 

STEPHENS,  GROVER  C.,  MARVA  J.  VOLK,  STEPHEN  H.  WRIGHT,  AND  PETER 

S.  BACKLUND 

Transepidermal  accumulation  of  naturally  occurring  amino  acids  in 

the  sand  dollar,  Dendraster  excentricus 335 

WURSIG,  BERND 

Occurrence  and  group  organization  of  Atlantic  bottlenose  porpoises 
(Tursiops  truncatus)  in  an  Argentine  Bay 348 

No.  3,  JUNE,  1978 

BOUSFIELD,  J.   D. 

Rheotaxis  and  chemoreception  in  the  freshwater  snail  Biomplmlaria 
glabrata  (Say):  estimation  of  the  molecular  weights  of  active  factors.  .  361 

DOERING,  G.   N.  AND  E.  E.   PALINCSAR 

Acid  phosphatase  during  the  life  cycle  of  the  nematode,  Panagrellus 
silusiae.  374 


CONTENTS  v 

FACTOR,  JAN  ROBERT 

Morphology  of  the  mouthparts  of  larval  lobsters,  Homarus  americanus 
(Decapoda:  Nephropidae) ,  with  special  emphasis  on  their  setae.  .  .  .  383 

FKLDER,  DARRYL  L. 

Osmotic  and  ionic  regulation  in  several  western  Atlantic  rallianassidae 
(Crustacea,  Decapoda,  Thalassinidea) 409 

HARRIGAN,  JUNE  F.  AND  DANIEL  L.  ALKON 

Larval  rearing,  metamorphosis,  growth  and  reproduction  of  the 
eolid  nudihranch,  Hermissenda  crassicornis  (Eschscholtz,  1831) 
(Gastropoda  :  Opisthobranchia) 430 

MORSE,  DANIEL  E.,  MARK  KAYNE,  MARK  TIDYMAN,  AND  SHANE  ANDERSON 
Capacity  for  biosynthesis  of  prostaglandin-related  compounds: 
distribution  and  properties  of  the  rate-limiting  enzyme  in  hydro- 
corals,  gorgonians,  and  other  coelenterates  of  the  Caribbean  and 
Pacific 440 

NAKAUCHI,  MITSUAKI  AND  KAZUO  KAWAMURA 

Additional  experiments  on  the  behavior  of  buds  in  the  ascidian, 
Aplidium  multiplicatum 453 

PERRON,  FRANK  E. 

Seasonal  burrowing  behavior  and  ecology  of  Aporrhais  occidentalis 
(Gastropoda :  Strombacea) 463 

RONAN,  THOMAS  E.,  JR. 

Food-resources   and    the    influence   of   spatial    pattern    on    feeding    in 

the  phoronid  Phoronopsis  viridis 472 

SASSAMAN,  CLAY  AND  JOHN  T.  REES 

The  life  cycle  of   Corymorpha    (  =  Euph ysora)   bigelowi   (Maas,    1905) 

and  its  significance  in  the  systematics  of  corymorphid  hydromedusae     485 

STEINACKER,  A. 

The  anatomy  of  the  decapod  crustacean  auxiliary  heart 497 

THORSON,  THOMAS  B.,  ROBERT  M.  WOTTON,  AND  TODD  A.  GEORGI 

Rectal  gland  of  freshwater  stingrays,  Potamotrygon  spp.  (Chondri- 
chthyes  :  Potamotrygon idae) 508 


Volume  154  Number  1 


THE 


BIOLOGICAL  BULLETIN 


PUBLISHED  BY 
THE  MARINE  BIOLOGICAL  LABORATORY 

Editorial  Board 

EDWARD  M.  BERGER,  Dartmouth  College  MEREDITH  L.  JONES,  Smithsonian  Institution 

JOHN  M.  ANDERSON,  Cornell  University  HOWARD  A.  SCHNEIDERMAN,  University  of 

California,  Irvine 
JOHN  B.  BUCK,  National  Institutes  of  Health 

RALPH  I.  SMITH,  University  of  California, 

JOHN  D.  COSTLOW,  Duke  University  Berkeley 

,  _  F.  JOHN  VERNBERG,  University  of 

PHILIP  B.  DUNHAM,  Syracuse  Umversity  South  Carolina 

J.  B.  JENNINGS,  University  of  Leeds  CARROLL  M.  WILLIAMS,  Harvard  University 

W.  D.  RUSSELL-HUNTER,  Syracuse  University 
Managing  Editor 


FEBRUARY,  1978 


'  • 

6 


Printed  and  Issued  by 
LANCASTER  PRESS,  Inc. 

PRINCE  &  LEMON  STS. 
LANCASTER,  PA. 


THE  BIOLOGICAL  BULLETIN 

THE  BIOLOGICAL  BULLETIN  is  issued  six  times  a  year  at  the  Lancaster  Press,  Inc.,  Prince  and 
Lemon  Streets,  Lancaster,  Pennsylvania. 

Subscriptions  and  similar  matter  should  be  addressed  to  THE  BIOLOGICAL  BULLETIN,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts.  Agent  for  Great  Britain:  Wheldon  and 
Wesley,  Limited,  2,  3  and  4  Arthur  Street,  New  Oxford  Street,  London,  W.  C.  2.  Single  numbers, 
$8.00.  Subscription  per  volume  (three  issues),  $22.00,  (this  is  $44.00  per  year  for  six  issues). 

Communications  relative  to  manuscripts  should  be  sent  to  Dr.  W.  D.  Russell- Hunter,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts  02543  between  June  1  and  September  1,  and 
to  Dr.  W.  D.  Russell-Hunter,  P.O.  Box  103,  University  Station,  Syracuse,  New  York  13210, 
during  the  remainder  of  the  year. 


Copyright  ©  1978,  by  the  Marine  Biological  Laboratory 
Second-class  postage  paid  at  Lancaster,  Pa. 


I   i    INSTRUCTIONS  TO  AUTHORS      - 

THE  BIOLOGICAL  BULLETIN  accepts  original  research  reports  of  intermediate  length  on  a  variety 
of  subjects  of  biological  interest.  In  general,  these  papers  are  either  of  particular  interest  to  workers 
at  the  Marine  Biological  Laboratory,  or  of  outstanding  general  significance  to  a  large  number  of 
biologists  throughout  the  world.  Normally,  review  papers  (except  those  written  at  the  specific 
invitation  of  the  Editorial  Board),  very  short  papers  (less  than  five  printed  pages),  preliminary 
notes,  and  papers  which  describe  only  a  new  technique  or  method  without  presenting  substantial 
quantities  of  data  resulting  from  the  use  of  the  new  method  cannot  be  accepted  for  publication.  A 
paper  will  usually  appear  within  four  months  of  the  date  of  its  acceptance. 

The  Editorial  Board  requests  that  manuscripts  conform  to  the  requirements  set  below; 
those  manuscripts  which  do  not  conform  will  be  returned  to  authors  for  correction  before  review 
by  the  Board. 

1.  Manuscripts.     Manuscripts  must  be  typed  in  double  spacing  (including  figure  legends, 
foot-notes,  bibliography,  etc.)  on  one  side  of  16-  or  20-lb.  bond  paper,  8}  by  11  inches.     They 
should  be  carefully  proof-read  before  being  submitted  and  all  typographical  errors  corrected 
legibly  in  black  ink.     Pages  should  be  numbered.      A  left-hand  margin  of  at  least  Ij  inches 
should  be  allowed. 

2.  Tables,  Foot-Notes,  Figure  Legends,  etc.    Tables  should  be  typed  on  separate  sheets  and 
placed  after  the  Literature  Cited.     Because  of  the  high  cost  of  setting  such  material  in  type, 
authors  are  earnestly  requested  to  limit  tabular  material  as  much  as  possible.     Similarly,  foot- 
notes to  tables  should  be  avoided  wherever  possible.     If  they  are  essential,  they  should  be  indi- 
cated by  asterisks,  daggers,  etc.,  rather  than  by  numbers.     Foot-notes  are  not  normally  permitted 
in  the  body  of  the  text.     Such  material  should  be  incorporated  into  the  text  where  appropriate. 
Explanations  of  figures  should  be  typed  double-spaced  and  placed  on  separate  sheets  at  the  end 
of  the  paper. 

3.  A  condensed  title  or  running  head  of  no  more  than  35  letters  and  spaces  should  be  included. 

Continued  on  Cover  Three 


It  is  with  deep  regret  that  THE  BIOLOGIC'. \i,  Kn.i.KTix  records 
the  death,  at  his  home  in  \Yoods  Hole  on  February  6.  1978,  of  Dr. 
Donald  P.  Costello,  Kenan  Professor  of  Zoology  Emeritus  of  the 
University  of  North  Carolina.  Chapel  Hill,  and  our  Managing 
Editor  from  1951  to  1968.  \Ye  know  that  many  readers  of.  authors 
in,  and  editorial  reviewers  for  TIIK  BIOLOGICAL  BULLETIN  ap- 
preciate the  extent  to  which  the  continued  international  standing  of 
the  journal  during  those  36  volumes,  and  even  subsequently,  de- 
pended upon  his  scholarly  efforts. 


Vol.  154,  No.  1  February,  1978 

THE 

BIOLOGICAL  BULLETIN 

PUBLISHED   BY   THE    MARINE   BIOLOGICAL  LABORATORY 


STUDIES  OX  FUNCTIONAL  MORPHOLOGY  IN  THE  DIGESTIVE 
SYSTEM  OF  OREASTER  RETICCLATL'S  (L.)  (ASTEROIDEA) 

Reference:  Biol.  Bull.,  154:  1-14.    (February,  1978) 

JOHN  MAXWELL  ANDERSON 

Division  of  Biological  Sciences,  Cornell  University,  Ithaca,  New  York  14853 

Orcastcr  reticulatus  (Fig.  1  )  is  a  large,  conspicuous,  and  easily-recognized  sea- 
star,  abundant  within  its  broad  geographical  range  and  probably  of  considerable 
ecological  significance  where  it  occurs.  Curiously,  however,  the  number  of  pub- 
lished papers  dealing  with  aspects  of  the  basic  biology  of  this  species  is  small.  A 
brief  note  by  Thomas  (1960)  describes  its  feeding  habits;  another,  almost  as  brief, 
by  Matthews  and  Lima-Verde  (1968)  suggests  an  ecological  relationship  between 
Orcastcr  and  two  species  of  Panulints  on  the  fishing-banks  of  Northeastern  Brazil. 
The  entire  literature  dealing  with  the  internal  anatomy  of  O.  reticulatus  appears  to 
consist  of  a  four-page  paper  with  three  plates  by  Tennent  and  Keiller  (1911  )  and  a 
brief  abstract  by  Anderson  (1967)  reporting  preliminary  observations  on  the 
digestive  system.  Some  special  features  of  the  internal  anatomy  of  Culcita,  a  genus 
assigned  to  the  Family  Oreasteridae,  were  described  and  figured  (rather  sketchily) 
by  Miiller  and  Troschel  (1842).  whose  drawing  was  reproduced  by  such  later 
authors  as  Luclwig  and  Hamann  (1899)  ;  but  beyond  this,  published  information 
on  Greasier  and  its  relatives  is  scanty.  The  purpose  of  the  present  paper  is  to 
describe,  in  greater  detail  than  that  provided  in  previous  accounts,  the  general 
morphology  of  the  digestive  system  as  a  whole,  noting  particularly  some  interesting- 
features  of  the  pyloric  stomach  and  related  structures.  This  is  intended  as  a  con- 
tribution toward  a  broadly  comparative  survey  of  digestive  systems  among  asteroids. 
Such  a  survey  has  never  been  undertaken  ;  a  step  in  this  direction  is  provided  by 
Anderson  (1966),  and  a  further  contribution  appears  in  Jangoux  and  van  Impe 
(1971),  but  available  data  are  still  inadequate  to  permit  comparisons  in  detail 
among  representatives  of  many  different  families  of  sea-stars.  Further,  an  attempt 
will  be  made  to  draw  together  published  and  unpublished  descriptions  of  feeding- 
behavior  in  Orcastcr.  and  to  correlate  these  with  structural  features  of  the  digestive 
system.  The  results  of  histological  studies  on  this  system  are  to  be  published  in  a 
subsequent  paper. 

Paraphrasing  a  statement  concerning  Orcastcr  that  he  had  published  in   1902, 

1 

Copyright  ©  1978,  by  the  Marine  Biological  Laboratory 
Library  of  Congress  Card  No.  A38-518 


JOHN  MAXWELL  ANDERSON 


FIGURE  1.     Greasier  retienlahts.     The  major  radius  of  this  specimen  measured  al>out   \2  cm 
(from  a  color  transparency  by  C.  P.  Anderson  I . 

FIGURE  2.     The  floor  of  the  cardiac  stomach,  internal   \ie\\.     The  arrow  indicates  the  dis- 


DIGESTIVE  SYSTEM  OF  OREASTER 

H.  L.  Clark  (1933,  p.  22)  says:  "This  is  undoubtedly  the  best  known  of  West 
Indian  sea-stars,  since  it  has  been  taken  to  many  parts  of  the  world  in  the  past  175 
years  as  a  typical  curio  and  souvenir  of  the  region."  It  is  a  widely  distributed 
species,  being  common,  according  to  Downey  (1973),  in  shallow-water  grass  and 
sand  Mats  from  Florida  to  Brazil;  it  occurs  in  the  Cape  Verde  Islands  and  in  Her- 
inuda,  and  it  is  occasionally  found  as  far  north  as  Cape  Hatteras.  Overall,  how- 
ever, the  scientific  interest  it  has  attracted  is  rather  limited.  A.  Agassiz  (  1877) 
gives  a  meticulous  and  beautifully  illustrated  description  of  its  skeletal  morphology, 
and  the  species  is,  of  course,  treated  in  systematic  and  faunistic  studies,  such  as 
those  cited  above,  and  others  (among  them  Verrill.  1915.  and  Caso.  1961).  In- 
creased attention  to  other  interesting  features  of  this  species  seems  long  overdue. 

MATERIALS  AND  METHODS 

Specimens  were  obtained  from  a  commercial  source  in  the  Florida  Keys.  At 
the  laboratory,  they  were  maintained  in  large  aquaria  provided  with  running  sea 
water  and  were  fed  periodically  with  crushed  snails,  shucked  bivalves,  and  pieces  of 
rish,  all  of  which  they  ate.  Under  these  conditions  they  remained  in  an  evidently 
healthy,  vigorous  state  until  sacrificed  for  study.  Animals  to  be  dissected  were 
first  soaked  until  flaccid  in  a  solution  of  MgSO4  ( 8r ;  in  tap  water).  The  speci- 
mens were  all  rather  large,  with  major  radii  ranging  from  10  to  14  cm.  In  such 
animals,  as  pointed  out  by  Tennent  and  Keiller  (  191 1  ),  the  body  wall  is  exceedingly 
hard  and  tough,  and  gaining  access  to  the  internal  organs  is  unusually  difficult. 
The  technique  eventually  adopted  was  as  follows:  using  a  strong,  sharp  scalpel 
(with  a  well-taped  handle),  a  horizontal  incision  was  made  around  the  periphery  of 

tinct  line  bounding-  tin-  smooth,  yellowish  central  area.  Abbreviation*  are:  in,  mouth  opening; 
/,  floor ;  and  />,  a  lateral  pouch. 

FIGURE  3.  Pouches  of  the  cardiac  stomach,  partially  everted  and  not  fully  inflated  with 
coelomic  fluid.  These  show  the  typical  asteroid  branching  gutter-patterns  in  the  wall,  here 
seen  from  the  mucosal  side. 

FIGURE  4.  Oral  view  of  the  origins  of  a  pair  of  Tiedemann's  ducts,  and  related  structures. 
Abbreviations  are:  </,  fibrous  girdle,  just  above  remnants  of  the  cut  wall  of  the  cardiac  stomach; 
r.  vertical  anchoring  strands  suspending  the  girdle  from  the  roof  of  the  disc;  Td,  Tiedemann's 
duct ;  and  og,  an  oral  gutter  of  Tiedemann's  pouch. 

FIGURE  5.  Detail  of  a  portion  of  a  pyloric  caecum,  with  associated  structures,  seen  from 
below.  Abbreviations  are:  og,  oral  gutter;  7'f>,  striated  side  wall  of  the  main  Tiedemann's 
pouch;  aT[>,  accessory  Tiedemann's  pouch;  and  Id,  lateral  diverticulum  of  a  glandular  pocket  of 
the  pyloric  caecum. 

FIGURE  6.  Oral  view  of  the  proximal  portions  of  two  radial  branches  of  the  pyloric  stomach, 
with  related  structures.  Abbreviations  are:  7V.  Tiedemann's  duct,  transected  at  .r;  '/"/>,  Tiede- 
mann's pouch;  rr,  radial  reservoir  of  the  pyloric  stomach;  and  li.\\  horizontal  connective-tissue 
sheet,  which  joins  with  the  vertical  anchoring  strands  to  provide  support  for  the  roof  of  the 
pyloric  stomach. 

FIGURE  7.  Aboral  view  of  the  roof  of  the  pyloric  stomach,  after  removal  of  the  intestine, 
intestinal  caeca,  and  rectum.  Abbreviations  are  :  o,  opening  from  the  pyloric  stomach  into  the 
intestine;  rf,  roof  folds  in  one  of  the  paired  fold-pattern  sets;  and  ind,  main  duct  or  channel  of 
a  fold-pattern,  showing  supposed  hemal  vessels  in  its  roof. 

FIGURE  8.  A  closer  view,  showing  structural  details  of  a  pair  of  fold-patterns  in  the  roof 
of  the  stomach. 

For  each  figure,  the  scale  bar  shown  represents  1  mm. 


4  JOHN   MAXWELL  ANDERSON 

each  rav,  and  the  incisions  in  adjacent  rays  were  joined  by  cutting  through  inter- 
radial  structures.  \Yhen  the  oral  and  aboral  parts  of  the  body  wall  were  thus 
separated,  the  aboral  wall  was  removed  by  carefully  cutting  all  mesenteries  and  con- 
nective-tissue strands  by  which  the  viscera  were  suspended  from  it,  and  by  transect- 
ing the  gut  just  below  the  anus.  The  viscera  were  then  floated  up  in  the  relaxing- 
solution,  the  dissection  continued,  and  the  desired  observations  made.  Anatomical 
details  were  photographed  in  freshly  dissected,  living  specimens,  using  a  35  mm 
camera  mounted  on  a  dissecting  microscope. 

OBSERVATIONS  AND  RESULTS 

The  central  cavity  of  the  cardiac  stomach  has  a  smooth  floor,  continuous  at  the 
mouth  with  the  external  peristomial  membrane  (Fig.  2).  The  floor  shows  fine 
radial  striations  or  wrinkles,  and  in  all  specimens  examined  it  was  of  a  yellowish 
color,  contrasting  with  the  generally  brown  cast  of  the  rest  of  the  stomach.  A  dis- 
tinct boundary-line  marks  the  margin  of  the  floor,  and  just  beyond  it  are  ranged 
the  regularly-spaced  oral  terminations  of  a  large  series  of  typical  branching  gutter- 
patterns  in  the  wall  of  the  stomach  (Fig.  3).  The  wall  is  folded  into  a  series  of 
pouches,  one  set  for  each  ray.  In  the  fully  retracted  condition  a  set  of  these  pouches 
lies  above  the  proximal  part  of  the  ambulacra!  ridge  pertaining  to  its  radius  in  the 
floor  of  the  disc,  bounded  on  either  side  by  an  interradial  septum.  Each  set  of 
pouches  comprises  a  thin-roofed  median  portion  and  a  pair  of  complexly  folded 
lateral  pockets.  The  lateral  pockets  or  pouches  in  adjacent  rays  are  joined  together 
by  smooth  walls,  thrown  into  loose  folds  and  passing  around  the  interradial  septa. 
These  smooth  walls,  with  inward  extensions  from  the  roofs  of  the  several  median 
pouches,  continue  upward  to  the  aboral  boundary  of  the  cardiac  portion  of  the 
stomach.  This  is  more  or  less  arbitrarily  recognized  as  a  slightly  constricted  region 
which  in  Oreastcr,  as  in  Patiria  (Anderson,  1959),  is  marked  by  an  encircling  con- 
nective-tissue girdle  related  to  the  suspensor-retractor  system  to  be  described  later. 
The  overlying  pyloric  stomach  is  tall  in  its  oral-aboral  dimension  and  is  deeply 
divided  radially  in  relation  to  the  origins  of  the  paired  pyloric  caeca  to  which  it 
gives  rise.  The  radial  indentations,  alternating  with  the  pyloric  caeca  and  their 
specialized  appendages,  are  like  those  described  in  several  other  sea-stars.  The 
similar  condition  found  by  Jangoux  and  van  Impe  (  1971  )  in  Astcrina,  Henricia, 
and  Echinaster  leads  these  authors  to  state  that  there  is  no  pyloric  stomach  in  the 
classical  sense,  but  rather  a  characteristic  structure  which  they  call  the  "  complex  c 
stomacal  superieure."  For  each  pyloric  caecum  a  separate  duct  originates  from  the 
stomach,  arising  just  above  the  level  of  the  girdle  (Fig.  4)  ;  using  the  terminology 
applied  to  the  similar  structure  in  Henricia  (Anderson,  1960,  p.  377),  this  is  re- 
ferred to  as  Tiedemann's  duct.  In  its  proximal  part  each  duct  is  a  cylindrical  tube, 
the  lumen  of  which  is  set  off  by  a  distinct  partition  from  the  space  above  it.  After 
proceeding  a  short  distance,  however,  it  opens  out  to  form  the  oral  gutter  of  a  deep 
Tiedemann's  pouch,  the  cavity  of  which  is  continuous  aborally  with  the  central 
duct  of  the  pyloric  caecum  proper.  This  central,  aboral  duct  (which,  it  will  be 
understood,  has  no  floor)  gives  off,  alternately  to  right  and  left,  side  branches 
which  are  the  ducts  of  a  long  series  of  typical  glandular  pockets,  extending  all  the 
way  to  the  distal  end  of  the  caecum.  The  side  walls  of  these  pockets  are  thickened 


DIGESTIVE  SYSTEM  OF  OREASTER  5 

and  folded  vertically  to  form  lateral  diverticula.  The  most  striking  and  con- 
spicuous feature  of  these  organs  is  the  presence  of  an  accessory  or  secondary  pouch 
hanging  below  each  lateral  glandular  pocket,  in  effect  branching  from  the  side  walls 
of  the  main  Tiedemann's  pouch.  They  are  provided  with  oral  gutters  which  origin- 
ate just  above  the  main  gutter,  and  they  taper  upward  and  outward  to  end  a 
variable  distance  from  the  tip  of  each  pocket.  The  walls  of  the  secondary  pouches, 
are  faintly  marked  by  narrow,  evenly-spaced,  parallel  vertical  striations  which  give 
the  appearance  of  separating  adjacent  channels  in  the  lumen  of  the  pouch.  Such 
markings  are  present  also  in  the  side  walls  of  the  main  Tiedemann's  pouch.  All 
of  these  features  are  shown  in  Figure  5.  In  all  respects,  the  pyloric  caeca  and 
Tiedemann's  pouches  of  0.  rcticnlatus  resemble  very  closely  the  corresponding- 
structures  in  Porania  pulrillns.  as  previously  described  (Anderson,  1961,  1966). 

It  was  mentioned  earlier  that  the  pyloric  stomach  is  deeply  indented  between 
the  bases  of  the  pyloric  caeca.  Alternating  with  the  indentations  are  radiating 
branches  of  the  stomach,  corresponding  in  position  with  the  Tiedemann's  ducts  just 
below  them;  their  side  walls  taper  outward  and  become  continuous  with  the 
proximal  ends  of  the  main  Tiedemann's  pouches.  Just  proximal  to  the  point  of 
origin  of  the  first  of  the  lateral  glandular  pockets  from  the  central  aboral  duct,  at 
about  the  level  where  the  tubular  Tiedemann's  duct  opens  to  become  the  oral 
gutter,  a  change  in  the  gross  appearance  of  the  side  walls  of  the  organ  occurs.  The 
line  marking  this  change  is  interpreted  as  the  boundary  between  Tiedemann's 
pouch  and  the  radial  branch  of  the  pyloric  stomach.  Aborally.  this  branch  receives 
the  central  duct  of  the  pyloric  caecum.  In  position  and  anatomical  relationships 
(Fig.  6),  it  corresponds  precisely  to  the  structures  observed  in  the  digestive  sys- 
tems of  PI  en  ri  cm  and  Linck'm  and  termed  "radial  reservoirs"  of  the  pyloric  stomach 
(Anderson,  1960,  1966). 

The  roof  of  the  pyloric  stomach  in  Orccistcr  presents  a  specialized  feature  which 
has  not.  to  my  knowledge,  been  described  in  any  other  sea-star.  Tennent  and 
Keiller  (1911)  write  of  it  as  follows  (p.  114)  :  "Beneath  the  intestine,  upon  the 
surface  of  the  stomach  in  each  radius,  is  what  appears  at  first  sight  to  be  a  second 
set  of  five  caeca,  each  made  up  of  two  parts.  Further  examination  shows  that 
these  are  merely  pouches  formed  by  the  folding  of  the  upper  wall  of  the  pyloric 
portion  of  the  stomach.  They  involve  the  regions  into  which  the  ducts  of  the 
pyloric  caeca  open  and  have  a  narrow  slit-like  connection  with  the  stomach."  Un- 
fortunately, these  unique  structures,  so  succinctly  characterized,  cannot  be  made 
out  with  certainty  in  Tennent  and  Keiller's  plate  showing  an  aboral  view  of  the 
digestive  system.  As  seen  in  Figure  7,  the  roof  of  the  pyloric  stomach  has  a  rela- 
tively smooth  portion  surrounding  the  opening  into  the  overlying  intestine.  Radiat- 
ing from  this  central  area  are  the  five  sets  of  "pouches"  just  described.  This  term 
seems  inappropriate ;  the  structures  referred  to  are  radially-arranged  fold-patterns, 
like  inverted  grooves  or  gutters,  in  the  roof  of  the  pyloric  stomach.  In  each  pattern, 
the  folds  converge  on  a  major  channel  leading  toward  the  intestinal  opening.  Each 
member  of  a  pair  of  fold-patterns  lies  above  a  radial  reservoir,  and  the  cavities  of 
the  fold-pattern  and  the  reservoir  communicate  by  way  of  the  "narrow  slit-like 
connections"  mentioned  by  Tennent  and  Keiller.  There  are  numerous  conspicuous 
vessels,  probably  parts  of  the  hemal  system,  running  in  the  aboral  walls  of  the  main 
channels  (Figs.  7,  8).  Figure  9,  a  semidiagrammatic  cross-section,  shows  the 


JOHN   MAXWELL  ANDERSON 


ROOF    FOLDS 


RADIAL 
RESERVOIR 
HORIZONTAL 
SHEET 


DUCT 


FIGURE  9.  Semidiagrammatic  cross-section  of  a  radial  portion  of  the  pyloric  stomach, 
showing  the  relationships  between  its  component  parts.  This  section  is  at  a  level  proximal  to 
the  point  at  which  Tiedemann's  duct  opens  out  to  form  the  oral  gutter  of  Tiedemann's  pouch. 
The  seam  forming  the  partition  between  the  duct  and  the  overlying  radial  reservoir  is  main- 
tained by  permanent  adhesion  between  cells  in  opposite  walls.  The  figure  was  made  by  tracing. 
\\ith  some  reconstruction,  a  projected  histological  section. 

relationships  between  the  fold-pattern,  the  radial  reservoir,  and  the  Tiedemann's 
duct  pertaining  to  a  single  pyloric  caecum.  The  close  correspondence  in  position 
between  the  paired  sets  of  fold-patterns  and  the  paired  pyloric  caeca  strongly  sug- 
gests a  significant  functional  relationship. 

The  intestine  in  Orcustcr  is  a  large,  flattened,  generally  pentagonal  organ.  From 
its  corners  five  flat  prolongations  extend  outward  in  the  interradii,  crossing  the 
roof  of  the  pvloric  stomach  between  adjacent  sets  of  fold-patterns  (Fig.  10).  Each 
extension  bifurcates  as  it  reaches  the  interradial  septum,  and  the  ten  branches  thus 
formed  become  the  very  large  and  conspicuous  intestinal  caeca.  The  main  duct 
gives  off  numerous  bladder-like  diverticula,  and  the  whole  organ,  according  to 
Tennent  and  Keiller,  is  capable  of  great  distention.  The  intestinal  caeca  of 
Oreastcr  arc  very  similar  to  those  of  Cnlcita.  as  illustrated  by  Miiller  and  Troschel 
(1842)  ;  see  also  Ludwig  and  Hamanu  (  1899,  p.  585  and  Plate  IV). 

The  rectum  i.s  short,  arising  from  about  the  center  of  the  roof  of  the  intestine 
and  passing  directly  through  the  aboral  body  wall.  The  opening  from  the  intestine 
into  the  rectum  lies  immediately  above  the  passage  leading  from  the  pyloric  stomach 
into  the  intestine  (Fig.  11  ). 

The  complement  of  fibrous  strands,  sheets,  and  mesenteries  developed  in 
Oreastcr  to  suspend  and  secure  the  digestive  organs  in  the  spacious  body  cavity, 
and  to  bring  about  retraction  of  the  eversible  parts,  is  very  complex.  Some  ac- 
count of  this  system  will  be  helpful  in  understanding  functional  relationships. 

The  pyloric  caeca  are  suspended  from  the  aboral  body  wall  by  the  usual  paired, 
parallel  mesenteries,  which  form  continuous  narrow  sheets  and  enclose  between 
them  a  long,  tubular  coelomic  space  above  the  central  duct  of  each  caecum.  The 
mesenteries  are  unusuallv  thick  and  tough  in  Oreastcr.  and  thev  send  short  ex- 


DKiKSTIVE  SYSTEM  OF   ORli. \STRR  7 

tensions  laterally  to  suspend  the  glandular  pockets  of  the  caecum.  Proximally,  the 
mesenteries  mingle  with  the  fibrous  bands  by  means  of  which  the  ducts  of  the 
intestinal  caeca  are  hung  from  the  roof  of  the  disc,  and  become  continuous  also 
with  a  thick  horizontal  sheet  which  covers  the  roof  of  the  pyloric  stomach  (Fig. 
10).  The  sheet  extends  between  the  proximal  ducts  of  the  intestinal  caeca,  and  it 
surrounds  and  attaches  to  the  lower  margins  of  the  paired  fold-patterns  in  the 
roof  of  the  pyloric  stomach.  In  each  of  the  paired  units,  the  level  of  attachment  to 
the  stomach  is  the  line  of  transition  between  fold-pattern  and  radial  reservoir 
(Figs.  6,  9.  10). 

Additional  suspension  is  provided  for  the  digestive  system  by  a  pair  of  bands 
in  each  rav  which  originate  on  the  aboral  body  wall  and  pass  downward.  Each  of 
these  vertical  bands  makes  a  connection  with  the  mesentery  complex  at  the  edge 
of  the  horizontal  sheet  and  then  continues,  passing  lateral  to  a  radial  reservoir  and 
joining  the  fibrous  girdle  encircling  the  cardiac  stomach  (Fig.  4).  Strong  bands, 
which  I  have  termed  "oral  anchors,"  proceed  from  these  junctions  on  the  girdle  to 
firm  attachments  alongside  the  proximal  ambulacral  ossicle  in  each  ray  (Figs.  12, 
13). 

In  each  radius,  a  group  of  glistening  white  extrinsic  retractor  strands  arises 
from  each  side  of  the  proximal  end  of  the  ambulacral  ridge.  From  broad  origins 
along  the  ridge,  the  fibers  converge  as  they  pass  upward  beside  the  pouches  of  the 
cardiac  stomach  and  form  three  major  branches  (Fig.  13).  One  of  these  spreads 
over  the  roof  of  the  median  pouch,  attaches  to  it  along  a  line  of  insertion  running 
radially,  and  sends  further  branches  orally  in  its  wall.  The  second  distributes 
principally  to  the  nearbv  lateral  stomach  pouch,  where  it  bifurcates  repeatedly 
(Fig.  14).  Its  branches,  and  those  of  the  first  major  strand,  give  rise  to  the 
downward-coursing  intrinsic  retractor  elements  on  and  in  the  walls  of  the  stomach, 
spreading  out  in  patterns  corresponding  to  those  of  the  gutters  mentioned  earlier. 
These  intrinsic  strands  are  very  similar  in  appearance  and  distribution  to  those 
designated  "class  1"  fillers  in  Patiria  (Anderson,  1959  i.  The  third  major  ex- 
trinsic branch  sends  a  few  subsidiaries  to  nearby  pouches  and  then  passes  directly 
to  the  girdle  on  the  cardiac  stomach,  which  it  joins  near  the  point  of  attachment  of 
one  of  the  vertical  suspensory  bands  descending  from  the  roof  of  the  disc.  I  have 
designated  this  branch  the  "girdle  retractor." 

It  is  to  be  understood  that  all  of  the  extrinsic  retractor  elements  just  described 
are  paired ;  that  is,  in  each  ray  there  are  two  sets  of  the  three  main  retractor 
branches,  which  arise  and  di>tribute  symmetrically  on  either  side  of  the  axis  of  the 
ray. 

Turning  again  to  the  intrinsic  retractor  strands,  it  is  worth  noting  that  in 
addition  to  the  class  1  type  previously  described,  other  small  branches  are  present. 
Considerable  numbers  of  short,  slender  fibers  originate  on  the  folds  and  ridges  of 
the  lateral  pouches  and  run  horizontally,  outside  the  wall  of  the  stomach,  before 
entering  it  again  (Fig.  15  i.  These  strongly  resemble  the  "class  2"  fibers  of 
Patina.  In  the  lowest  part  of  the  stomach  groups  of  3  to  12  thin  strands  emerge 
from  folds  related  to  the  terminal  gutter-patterns  and  run  vertically,  free  in  the 
coelom.  to  insertions  on  the  outer  surface  of  the  smooth  floor  of  the  cardiac  stomach 
(Figs.  1C).  17).  They  are  similar  to  the  "class  3"  fillers  found  in  Patina  (Ander- 
son, 1969).  Precise  correspondence  between  these  fiber  types  cannot  be  firmly 


8 


JOHN  MAXWKI.T.  ANDKKSON 


FIGURE  10.  Aboral  view  of  a  portion  of  the  roof  of  the  pyloric  stomach  showing  its  rela- 
tionship to  the  intestine  and  intestinal  caeca.  Abbreviations  are :  i,  intestine ;  </,  duct  leading  to 
a  pair  of  intestinal  caeca,  proximal  to  its  point  of  bifurcation  ;  /,  fold-pattern  ;  and  lis.  hori- 
zontal connective-tissue  sheet.  The  arrow  indicates  the  margin  of  the  duct  where  the  hori- 


DIGESTIVE  SYSTEM  OF  OREASTER  <> 

established,  however,  without  information  on  their  histological  characteristics.  In 
Patina,  class  2  fibers  are  muscular,  while  those  designated  class  3  appear  to  con- 
sist of  thin  strands  of  connective  tissue. 

DISCUSSION 

The  distinctive  combination  of  special  features  presented  by  the  digestive  system 
of  O.  reticulatns  suggests  that  this  sea-star  is  capable  of  considerable  versatility  in 
its  feeding  habits.  The  unusually  large,  extensively  eversible  cardiac  stomach,  with 
its  well-developed  systems  of  anchoring  and  retracting  fibers,  is  structurally  and 
probably  functionally  similar  to  that  of  Patina  (Anderson,  1959).  It  seems 
primarily  adapted  for  handling  large  pieces  of  food  outside  the  body,  in  the  manner 
characteristic  of  many  carnivorous  or  omnivorous  sea-stars.  Observations  made  in 
the  course  of  the  present  study,  on  specimens  maintained  in  aquaria,  confirm  that 
Orcastcr  does  envelop  food  in  everted  folds  of  the  cardiac  stomach.  In  feeding  on 
a  piece  of  fish,  for  example,  the  animal  first  dilates  its  mouth;  several  flattened, 
somewhat  palmate  lobes  of  the  stomach  (probably  the  lateral  pouches  described 
earlier)  protrude  in  contact  with  the  food  and  then  surround  it  as  they  are  further 

zontal  sheet  attaches  and  binds  it  down  to  the  roof  of  the  stomach.  Above  the  arrow  may  be 
seen  three  of  the  many  strands  by  which  the  duct  is  suspended  from  the  roof  of  the  disc. 

FIGURE  11.  Central  portion  of  the  roof  of  the  pyloric  stomach,  as  viewed  from  the  oral 
side.  Note  the  coarse  folds  (/)  that  hang  down  between  the  slit-like  openings  from  the  paired 
sets  of  fold-patterns.  Other  abbreviations  are :  rps,  roof  of  the  pyloric  stomach ;  r  rectum ;  and 
;',  intestine.  The  anus  opens  immediately  above  the  short  rectum  seen  here. 

FK;URK  12.  Aboral  view  of  a  portion  of  the  fibrous  girdle  (</)  encircling  the  upper  part  of 
the  cardiac  stomach,  showing  two  of  the  ten  oral  anchors  (oa)  that  attach  it  to  the  proximal 
ambulacra!  ossicles,  and  a  girdle  retractor  (gr)  representing  part  of  the  extrinsic  retractor 
system  (see  Fig.  13). 

FIGURE  13.  One  of  the  paired  sets  of  extrinsic  retractor  strands  in  a  ray,  showing  the 
distribution  of  its  principal  branches.  Abbreviations  are:  cr,  the  main  extrinsic  retractor  near 
its  origin  alongside  the  ambulacral  ridge ;  ;/;/>,  the  branch  that  inserts  principally  on  the  roof  of 
the  median  pouch  of  the  cardiac  stomach ;  //>,  the  branch  that  turns  laterally  under  the  preceding 
one  and  distributes  to  a  lateral  pouch  of  the  cardiac  stomach  (see  Fig.  14)  ;  and  gr,  a  girdle 
retractor,  the  third  major  branch,  whose  stoutest  portion  runs  directly  to  the  girdle  and  at- 
taches there.  The  remaining  large  strand  (oa)  is  one  of  the  oral  anchors  of  the  girdle  (r/. 
Fig.  12). 

FIGURE  14.  External  view  of  a  portion  of  one  lateral  pouch  of  the  cardiac  stomach,  showing 
the  repeatedly  bifurcating  class  1  intrinsic  retractor  fibers  distributing  in  the  wall. 

FIGURE  15.  External  view  of  a  portion  of  a  lateral  pouch  photographed  after  fixation  in 
Bouin's  fluid  (to  provide  enhanced  contrast).  Some  downward-coursing  class  1  retractors  are 
shown  (cl  1),  as  well  as  a  number  of  the  slender  class  2  fibers  (cl  2)  that  stretch  horizontally 
between  adjacent  folds  of  the  stomach  wall. 

FIGURE  16.  The  floor  of  the  cardiac  stomach  viewed  from  the  coelomic  side;  specimen 
photographed  after  fixation  in  Bouin's  fluid.  Sets  of  slender  fibers  (cl  3)  are  seen  running 
vertically  from  the  gutter-patterns  to  attach  outside  the  smooth  floor  of  the  stomach.  These  are 
provisionally  identified  as  class  3  intrinsic  retractor  fibers. 

FIGURE  17.  Part  of  an  everted  vesicle  of  the  cardiac  stomach,  seen  from  its  mucosal  side. 
showing  part  of  the  array  of  terminal  branches  of  the  gutter-patterns  (<//>)  bordering  the 
smooth  floor  of  the  stomach  (/).  Through  the  thin  wall  of  the  vesicle  may  be  seen  some  of  the 
slender,  vertical,  parallel  strands  of  the  supposed  class  3  intrinsic  retractors,  attached  to  the 
coelomic  side  of  the  stomach  wall. 

For  each  figure,  the  scale  bar  shown  represents  1  mm. 


10  JOHN    MAXWELL   AXDKUSOX 

inflated  \vitli  coeloinic  fluid.  The  characteristic  branching  gutter-patterns,  with 
their  associated  intrinsic  retractor  strands,  can  he  clearly  seen  on  the  vesicles  of  the 
stomach.  The  animal  may  remain  with  its  stomach  everted  for  several  hours,  as 
the  food  gradually  disintegrates  and  the  products  of  digestion  are  transported  to  the 
inner  parts  of  the  system.  In  the  only  published  account  of  the  feeding  habits  of 
O.  rcticuhitits  under  natural  conditions,  Thomas  (1960)  reports  having  observed  a 
specimen  with  its  stomach  everted  over  a  small  unattached  sponge  which  appeared 
to  have  been  partially  digested.  Thomas  also  cites  an  unpublished  account  by 
another  observer  who  saw  Orcastcr  feeding  on  a  sponge.  Further  unpublished 
observations  hv  Dr.  (erald  I  lalpern  (communicated  to  me  by  letter)  confirm  the 
fact  that  O.  rcfic/i/a/its  consumes  large  pieces  of  detritus.  Toponce  (1973)  reports 
that  the  related  Eastern  Pacific  species  O.  occidentals  feeds  on  clumps  of  stony 
coral,  or  on  bits  of  algae.  All  available  evidence  thus  substantiates  the  supposition 
that  might  have  been  made  on  anatomical  grounds  alone:  that  Orcastcr  functions 
as  a  macrophagous  carnivore  or  scavenger. 

It  is  evidently  capable  of  other  modes  of  feeding  as  well.  Thomas  (1960) 
writes  :  "I  have  observed  Orcastcr  many  times  with  its  stomach  everted  into  small 
depressions  in  the  coralline  sand  and  Thalassia  bottoms  on  which  it  lives.  Examina- 
tion reveals  nothing  either  in  the  depression  or  in  the  stomach  which  might  be  of 
food  value.  Possibly  any  organic  material  close  to  the  stomach  wall  is  digested  in 
this  manner."  I  lalpern  (personal  communication)  also  mentions  having  observed 
this  type  of  feeding  behavior  in  Orcastcr.  The  phenomenon  is  interesting  in  view 
of  the  analogous  behavior  exhibited  by  I'atiria  ininiata,  as  described  by  Anderson 
(1959).  This  species  is  very  frequently  seen  in  tide  pools  with  its  voluminous 
cardiac  stomach  fully  everted,  although  no  visible  objects  of  food  are  enfolded  by  it. 
In  an  aquarium,  the  animal  applies  its  everted  stomach  to  the  wall,  as  though  di- 
gesting the  film  of  microorganisms  adhering  to  the  glass.  Anderson  (1959)  sug- 
gested that  Patiria  might  be  using  its  everted  stomach  as  a  flagellary-mucous  feed- 
ing organ,  to  collect  suspended  participate  matter.  Araki  (1964)  reports  that 
under  experimental  conditions  specimens  of  Patiria  with  everted  stomachs  are  cap- 
able of  rapidly  removing  organic  compounds  from  solution  in  the  surrounding  wa- 
ter, and  his  suggestion  is  that  the  stomach  is  involved  in  this  function.  Although 
there  appears  to  be  some  question  as  to  the  significance  of  dissolved  organic  matter 
in  the  overall  nutrition  of  marine  animals  (Jdrgensen,  1976),  the  thin-walled 
stomach  seems  ideally  suited  to  mediate  whatever  exchange  of  materials  may  take 
place  between  sea  water  and  the  enclosed  coelomic  fluid.  All  things  considered, 
one  may  justifiably  conclude  that  whatever  Patiria  is  doing  with  its  stomach 
everted  in  the  absence  of  visible,  macroscopic  food.  Orcastcr  is  probably  doing 
something  similar. 

further  evidence  of  versatility  in  feeding  is  provided  bv  the  presence  of  highly 
specialized  features  in  the  digestive  system  above  the  level  of  the  cardiac  stomach. 
These  include  the  very  elaborate  Tiedemann's  pouches,  the  highly  folded  structures 
in  the  roof  of  the  pyloric  stomach,  and  the  unusually  large  intestinal  caeca.  Such 
features  as  these  are  never  found  in  strictly  carnivorous  sea-stars  such  as  Aster ias 
and  its  relatives ;  there,  the  pyloric  stomach  is  small  and  simple,  Tiedemann's 
pouches  are  lacking,  and  the  intestinal  or  rectal  caeca  are  strongly  reduced.  The 


DKiKSTlVE   SYSTKM    OF   OKI-ASTER  11 

more  highly  specialized  structures  are  characteristic  of  forms  known  to  he.  or 
suspected  of  being,  microphagous  particle-feeders.  In  all  sea-stars,  even  carnivore^. 
there  is  a  consistent  pattern  of  flagellary  circulation  through  the  digestive  system. 
Tiedemann's  pouches  are  interpreted  as  flagellary  pumping  organs,  functioning  to 
enhance  the  volume  and  velocity  of  water-flow  through  the  gut  in  connection  with 
the  exploitation  of  suspended  participate  matter  as  food  (Anderson.  I960). 

The  remarkable  anatomical  similarity  between  the  Tiedemann'>  pouches  of  (>. 
rcticitlatits  and  those  of  P crania  is  significant  in  this  regard.  As  long  ago  as  1915. 
Gemmill  provided  experimental  evidence  that  Porania  can  be  maintained  without 
weight-loss  for  long  periods  (several  months)  with  no  food  other  than  suspended 
particles.  In  describing  Tiedemann's  pouches  in  Porania  many  years  later,  Ander- 
son (1961)  called  attention  to  their  unusually  complex  structure,  involving  the 
development  of  many  subsidiary  pouches  branching  from  the  main  one.  The  con- 
clusion is  unavoidable  that  this  elaboration  is  related  to  the  demonstrated  ability  of 
Porania  to  subsist  on  participate  food  alone.  It  is  of  interest  that  Jangoux  (1972) 
has  described  similar  secondary  Tiedemann's  pouches  in  Archastcr  ani/iilutus :  they 
are  present  also  in  Dermasterias  inihricaia  (Anderson,  unpublished  observations). 
It  is  perhaps  not  unreasonable  to  suggest  that  sea-stars  with  subsidiary  or  accessory 
Tiedemann's  pouches  are  at  least  facultative  particle-feeders. 

Species  in  which  Tiedemann's  pouches  are  well  developed  characteristically 
possess  much  larger  intestinal  caeca  than  those  lacking  such  structures.  Hcnncia 
and  Patiria  both  demonstrate  this  correlation  to  a  considerable  degree  (Anderson, 
1966),  as  do  their  relatives  lichinastcr  and  Asicrina  (  Jangoux  and  van  Impe,  ll)71  i. 
If  the  supposition  is  justified  that  Tiedemann's  pouches  are  significantly  related  to  a 
capacity  for  particle-feeding  (Anderson,  1960),  it  is  tempting  to  go  one  step  further 
and  suggest  that  well-developed  intestinal  caeca  are  also  involved  somehow  in  thi> 
function.  Here  again.  Porania  piilrillns  provides  a  key  example.  The  intestinal 
caeca  of  this  species  are  very  large  indeed,  and  (iemmill  (1^15.  p.  12)  describe> 
their  rhythmic  contraction  and  expansion,  "sometimes  with  such  activity  as  to 
suggest  the  systole  of  the  auricular  portion  of  a  heart."  According  to  Gemmill. 
Porania  periodically  inflates  its  gut  with  water,  which  is  drawn  in  through  the 
mouth  and  later  expelled  forcefully  from  the  anus.  ( iemmill  believed  that  the  large 
and  muscular  intestinal  caeca  are  responsible  for  the  expulsion.  Since  the  caeca 
lack  any  intrinsic  mechanism  for  expansion,  it  seems  likely  that  the  pressure  re- 
quired to  inflate  them  with  water  is  provided  by  the  large  Tiedemann's  pouches, 
whose  centripetal  currents  converge  on  the  roof  of  the  pyloric  stomach  and  enter 
the  intestine. 

The  remarkably  large,  well-developed  intestinal  caeca  of  Orcastcr  and  of  its 
relative  Culcita  have  been  referred  to  earlier,  and  it  will  be  recalled  that  Tennent 
and  Keiller  (1911)  described  the  intestinal  caeca  as  capable  of  great  distention. 
They  say  further  (p.  114):  "Upon  opening  some  specimens  the  caeca  were  found 
to  be  greatly  distended.  Upon  stimulation  they  slowly  contracted,  the  entire  organ 
shrinking  to  about  one-third  of  its  former  size.  The  contents  were  watery  .  .  ." 
Although  we  lack  for  Orcaster  any  comprehensive  series  of  observations  on  water- 
flow  through  the  gut,  and  on  filter-feeding,  such  as  those  provided  by  Gemmill 
(1915)  for  Porania,  the  structural  similarities  between  the  two  forms  strongly  sug- 


12  JOHN   MAXWELL  ANDERSON 

gest  comparable  functions  ;  and  direct  evidence  is  not  altogether  lacking.  Halpern, 
cm  the  basis  of  unpublished  observations,  is  convinced  that  O.  rcticulatns,  in  addition 
to  its  other  modes  of  nourishing  itself,  is  indeed  a  filter-feeder.  His  letter,  pre- 
viously referred  to,  states:  "In  the  area  I  observed  it  in,  it  filter-feeds  when  there 
is  a  strong  tidal  current.  There  are  many  loggerhead  sponges  (Spheciospongia 
rcsparia'),  and  Orcastcr  often  uses  these  as  a  purchase  so  as  to  be  able  to  outstretch 
one,  two,  or  even  three  arms.  As  the  current  slackens,  they  abandon  this  method 
of  feeding.  Both  the  fact  of  filter-feeding  and  the  current  acting  as  a  stimulus  have 
been  confirmed  (but  not  conclusively)  by  some  preliminary  laboratory  experi- 
ments." 

One  further  specialization  in  relation  to  water-movement  in  the  gut,  and  pos- 
sible particle-feeding,  is  represented  by  the  folded  structures  in  the  roof  of  the 
pyloric  stomach.  Abundant  wrinkles  in  this  general  area  are  known  in  a  number  of 
sea-stars;  Tiedemann  (1816)  showed  something  of  the  kind,  with  associated  vessels, 
in  Astropecten  mtrantiacus,  and  according  to  Jangoux.  Perpeet,  and  Cornet  (1972) 
such  structures  are  particularly  well-developed  in  Astcrias  ritbcns.  I  know  of  no 
other  species,  however,  in  which  strongly  flagellated,  radially  folded  patterns  have 
been  described,  lying  in  such  an  obviously  functional,  oriented  relationship  between 
the  radial  reservoirs  and  the  opening  from  the  pyloric  stomach  into  the  intestine,  as 
they  do  in  Orcastcr  rcticulatus.  ( I  am  informed  by  Dr.  Michel  Jangoux,  however, 
that  in  an  extensive  series  of  unpublished  observations  on  members  of  the  Family 
Oreasteridae  he  has  found  similar  structures  in  Pcntacerastcr,  Protoreastcr,  and 
Culcita).  In  a  dissected  specimen  of  Orcastcr,  very  rapid  currents  can  be  demon- 
strated, using  dilute  India  ink,  running  through  the  fold-patterns  from  the  central 
ducts  of  the  pyloric  caeca  toward  the  intestine.  In  my  interpretation,  the  fold- 
patterns  in  the  roof  of  the  pyloric  stomach,  the  unusually  large  intestinal  caeca, 
and  the  elaborate  Tiedemann's  pouches  form  a  coordinated  set  of  adaptations  which 
enable  Orcastcr  to  utilize  suspended  participate  matter  as  a  source  of  food.  Inter- 
estingly, all  of  these  features  are  present  in  other  members  of  the  Family  Orea- 
steridae (Jangoux,  Universite  Libre  de  Bruxelles,  personal  communication). 

It  is  clear,  however,  from  all  the  evidence,  morphological  as  well  as  behavioral, 
that  Orcastcr  is  not  exclusively,  or  perhaps  even  primarily,  a  particle-feeder,  as 
Hcnricia  may  be  (Anderson.  1960;  Rasmussen,  1965).  It  should  be  borne  in  mind 
that  even  that  celebrated,  demonstrated  particle-feeding  species,  Porania  puh'illus, 
internally  so  similar  to  Orcastcr,  does  not  depend  altogether  on  a  participate  diet. 
Gemmill's  statement  (1915,  p.  14)  that  "At  the  Millport  Marine  Station  the 
Porania  are  never  seen  feeding  on  shell-fish,  etc.,  or  on  their  neighbors  as  other 
species  readily  do"  seems  to  imply  some  such  conclusion;  but  he  goes  on  to  say 
only  that  "ciliary  feeding  plays  a  part  in  the  nutritional  economy  of  Porania." 
Recent  studies  by  Ericsson  and  Hanssen  (1973  )  have  shown  that  Porania  pulvillus, 
in  fact,  feeds  on  octocorals,  brachiopods,  and  ascidians,  both  in  its  natural  habitat 
and  in  aquaria. 

It  is  to  be  hoped  that  similar  studies  may  soon  be  made  on  the  feeding  biology  of 
Orcaster  rcticulatns,  to  supplement  incidental  observations,  and  to  determine  the 
validity  of  conclusions  that  can  now  only  be  inferred  from  consideration  of  the 
comparative  anatomy  of  the  digestive  system. 


DIGESTIVE  SYSTEM  OF  OREASTER  13 

The  dissections  and  observations  on  which  this  report  is  based,  together  with 
preliminary  histological  procedures,  were  carried  out  at  the  Mote  Marine  Labora- 
tory, Siesta  Key,  Sarasota,  Florida.  It  is  a  pleasure  to  express  to  the  Director  of 
the  Laboratory,  Dr.  Perry  \Y.  Gilbert,  my  appreciation  for  his  gracious  hospitality, 
for  excellent  facilities  generously  provided,  and  for  the  valuable  assistance  of  mem- 
bers of  the  staff,  particularly  Pat  Bird  and  Susi  Dudley.  I  am  grateful  also  to 
Dr.  Jerald  Halpern  and  Dr.  Michel  Jangoux  for  providing  details  of  their  unpub- 
lished observations  on  Oreastcr  and  its  relatives. 

SUMMARY 

This  paper  presents,  with  illustrations,  a  description  of  the  digestive  system  of 
Oreaster  reticnlatns,  a  species  for  which  such  anatomical  details  have  hitherto  been 
unavailable.  Special  features  of  the  digestive  system  include  a  large,  highly 
eversible  cardiac  stomach  with  a  particularly  well-developed  system  of  securing  and 
retracting  fibers  ;  a  highly  specialized  pyloric  stomach  giving  rise  to  paired  pyloric 
caeca,  each  of  which  is  associated  with  an  unusually  elaborate  Tiedemann's  pouch 
featuring  a  series  of  secondary  pouches  branching  off  along  its  length  ;  and  a  set 
of  very  voluminous  intestinal  caeca.  By  comparison  with  other  asteroids  for  which 
anatomical  details  and  feeding  biology  are  known  (especially  Patiria  ininiata  and 
Porania  puh'illus),  it  is  suggested  that  ().  rcticulatus  is  equipped  for  a  variety  of 
modes  of  feeding.  The  cardiac  stomach  is  well  adapted  for  the  digestion  of  large 
pieces  of  food  outside  the  body ;  it  may  also  function  as  a  flagellary-mucous  particle- 
collector,  as  the  similar  organ  of  Patiria  is  thought  to  do.  The  specializations  of 
the  upper  part  of  the  digestive  system  are  closely  similar  to  corresponding  organs 
in  the  known  particle-feeding  species  Porania  piih'illus,  and  it  seems  probable  that 
Oreastcr  may  use  its  Tiedemann's  pouches  and  intestinal  caeca  to  bring  particle- 
laden  water  into  the  digestive  system  in  a  manner  similar  to  that  described  for 
Porania.  Such  direct  observations  as  are  available  on  the  feeding  behavior  of  O. 
rcticulatus  tend  to  confirm  the  conclusions  inferred  on  indirect,  anatomical  grounds. 

LITERATURE  CITED 

AGASSIZ,  A.,  1877.     North  American  starfishes.     Mem.  Mus.  Comp.  Zoo/.  Harvard,  5:  1-136. 
ANDERSON,  J.  M.,  1959.     Studies  on  the  cardiac  stomach  of  a  star-fish,  Patiria  ininiata  (Brandt). 

Biol.Bull.,  117:  185-201. 
ANDERSON,  J.  M.,   1960.     Histological  studies  on  the  digestive  system  of  a   starfish,  Henricia, 

with  notes  on  Tiedemann's  pouches  in  starfishes.     Biol.  Bull.,  119:  371-398. 

ANDERSON,  J.  M.,  1961.     Structural  peculiarites  of  the  pyloric  caeca  in  a  particle-feeding  sea- 
star,  Porania  puh'illus.    Am.  Zoo/.,  1  :  338-339. 
ANDERSON,  J.  M.,  1966.    Aspects  of  nutritional  physiology.    Chapter  14  in  R.  A.  Boolootian,  Ed.r 

Physiolofiy  of  Ecliinodcrniata.     Wiley-Interscience,  New  York. 
ANDERSON,  J.  M.,  1967.     Some  details  of  the  digestive  system  in  a  sea-star,  Oreastcr  rcticulatus. 

Am.  Zoo/.,  7  :  770. 
ARAKI,  G.  S.,  1964.     On  the  physiology  of  feeding  and  digestion  in  the  sea  star  Patiria  ininiata. 

Ph.  D.  Dissertation ,  Stanford  Unh'crsitv,   194  pp.    (Diss.  Ahstr.,  25:   4306;   order  no. 

64-13,558). 
CASO,  M.  E.,  1961.     Estado  actual  de  los  conocimientos  acerca  de  los  Equinodermos  de  Mexico. 

Ph.  D.  Dissertation.   L'nircrsidad  Nacional  Autunonia  dc  Mexico,  Mexico,  D.   F.,  388 

pp. 
CLARK.  H.  L.,  1902.     The  echinoderms  of  Porto  Rico.    Bull.  U.  S.  Fish.  Coiiiin.  (for  1900},  20: 

231-263. 


14  JOHN    MAXWELL   AXDHRSOX 

CLARK,  H.  I...  1933.     ./  handbook  of  Ilic  littoral  echinoderms  <>f  I'orlo  J\icn  and  the  nllier  ll'est 

Indian  islands.     Scientific  Suri'cy  of  Porto   l\'ico  and   Ilic    I'iriiin   Islands,   16(1).   \i-\\ 

York  Academy  of  Sciences,  New  York,  147  pp. 
DOWNEY,  M.  E.,  1973.     Starfishes  from  the  Caribbean  and  the  ( iulf  of  Mexico.     Sinillisnn.  Con- 

trih.  Zool,  126:  1-158. 
KkirssoN,   J.,   A NII    H.    G.    HANSSKX,    1973.      Observations    on    the    feeding   biology   of   Porania 

pnk'illus   (O.  V.  Miiller),   (Asteroidea),  from  the  Swedish  West  Coast.     Ophelia,  12: 

53-58. 
GEMMII.I.,  J.  F.,   1915.     On  the  dilation  of  asterids,  and  on  the  question  of  ciliary  nutrition  in 

certain  species.     Proc.  Zool.  Soc.  Lond.,  1915  :   1-19. 
JANGOUX,  M.,   1972.     Note  anatomique  sur  Arcliaster  angulatus  Miiller  et  Troschel    (  Echino- 

dermata,  Asteroidea).     Kcr.  Zool.  Bot.  Ajr..  86:   163-172. 
JANGOUX,  M.,  AND  E.  VAN  IMPE,  1971.     fitude  comparative  des  activates  phosphomonesterasiques 

alcalines  du   tube   digestif  de   plusieurs   especes  d'asterokles    (  fichinodermes)    precedee 

d'une  note  anatomique.     Call,  liiol.  Mar.,  12:  405-418. 
JANGOUX,  M.,  C.  PERPEET,  AND  D.  CORNET,  1972.     Contribution  a  1'etude  des  poches  stomacales 

d'Asterias  nihens   (Echinodermata :  Asteroidea).     Mar.  Biol..  15:  329-335. 

J0RGENSEN,  C.  B.,  1976.     August  Putter,  August  Krogli,  and  modern  ideas  on  the  use  of  dis- 
solved organic  matter  in  aquatic  environments.     Biol.  AVr.,  51  :  291-328 
l.ritwiG,  H.,  AND  O.  HAMANN,  1899.     Echinodermen,  II.   Buch.   Die  Seesterne.     Pages  461-966 

in  H.  G.  Bronn,  Ed.,  Klassen   iind  Ordiiuni/en  des  Tlncrreiclis.  Bd.  2,  Aht.  3.  Winter, 

Leipzig. 
MATTHEWS,  H.  R.,  AND  J.  S.  LIMA-VERDE,  1968.     Notas  sobre  Orcaster  reticiilatns   (Linnaeus, 

1758)  no  nordeste  Brasileiro  (Echinodermata:    Asteroidea).    Arq.  Est.  Biol.  Mar.  L'uii: 

Fed.  Ceard,  8:  223-224. 

MULLER,  J.,  AND  F.  H.  TROSCHEL,  1842.    System  der  Asferiden.    A'ieweg,  Braunschweig.  134  pp. 
RASMUSSEN,   B.    X.,    1965.     On   taxonomy   and   biology  of   the   North   Atlantic    species   of   the 

Asteroid  Genus  Hcnricia  Gray.     Medd.  Dan.  Fisk.-Havsunders.,  4:   157-213. 
TENNENT,   D.   H.,   AND  V.   H.   KEILLER,   1911.     The  anatomy  of  Peutaeeros   reticiilatns.     l\ip. 

Tortitt/cis  Lab.,  Car)ie</ic  lust,  li'asli..  3:   113-116. 
THOMAS,   L.   P.,    1960.     A   note  on   the  feeding   habits   of  the   West   Indian   sea   star   Oreaster 

rcticuhitus  (Linnaeus).     Q.  J.  Fla.  Acad.  Set..  23:  l(>7-lo8. 
TIEDEMANN,  F.,   1816.     Anatomic  der  Rohrenholothurie  ties   pomeransfarbigen   Secstenis   mid 

Stein-Seeit/els.   Thotnann,  Landshut,  98  pp. 
TOPONCE,  D.,  1973.'    Cabo  Pulmo  Reef.     Oceans.  6  :  42-45. 
VEKKILL,  A.  E.,   1915.     Report  on  the  starfishes  of  the  West   Indies,   Florida,  and  Brazil.   Bull. 

Lab.  Nat.  Hist.  State  Univ.  loiva,  7  :  1-232. 


Reference:  liiol.  Bull.,  154:  15-.il.  (  February, 


INTERACTION  OF  IONIZED  AND  UN-IONIZED  AMMONIA  ON 

SHORT-TERM  SURVIVAL  AND  GROWTH  OF  PRAWN 

LARVAE,  MACROBRACHICM  ROSEN BERGII 

DAVID  A.  ARMSTRONG,  DEBBIE  CHIPPENDALE,  ALLEN  W.  KNIGHT 

AND  JOHN  E.  COLT 

H\drobiolog\  Laboratory,  Department  of  Land,  Air  and  Heater  Resources,  Water  Science  and 

Engineering  Section,  University  of  California,  Davis,  California  95616;  and  Department 

of  Civil  Engineering,  University  of  California,  Davis,  California  95616 

Ammonia  is  the  principal  excretory  product  of  Crustacea  (Hartenstein,  1970; 
Hochachka  and  Somero,  1973;  Kinne,  1976),  and  its  modes  of  toxicity  as  well  as 
concentrations  lethal  to  a  variety  of  organisms  have  been  well  documented  (Warren, 
1962;  Campbell,  1973).  Ammonia  exists  in  solution  primarily  as  the  NH4+  ion  and 
the  un-ionized  NH3  molecule,  the  proportions  of  which  are  highly  pH-dependent. 
In  this  paper  ammonia  will  refer  to  the  sum  of  NH4+  and  NH3.  Un-ionized  am- 
monia will  refer  to  the  NH8  molecule  and  ionised  ammonia  to  the  NH4+  form. 

In  the  aquatic  habitat,  organisms  rely  on  rapid  diffusion  of  NH3  across  the  gill 
membranes  (Fromm  and  Gillette,  1968)  or  exchange  transport  of  NH4+  with  Na+ 
(Maetz  and  Garcia-Romeu,  1964;  Campbell,  1973;  Mangum  and  Towle,  1977)  to 
void  themselves  of  this  toxicant.  Diffusion  of  NH3  is  a  principal  route  of  excretion 
because  blood  levels  are  normally  much  greater  than  ambient  concentrations  (see 
Kinne,  1976,  for  review).  Fromm  and  Gillette  (1968)  reported  that  ammonia 
levels  in  the  blood  of  trout  are  9-40  times  greater  than  in  ambient  wrater.  Concen- 
trations of  ammonia  in  the  blood  of  Crustacea  range  from  2  to  18  mg/liter  (Myers, 
1920;  Florkin  and  Renwart,  1939;  Mangum,  Silverthorn,  Harris,  Towle  and  Krall, 
1976),  which  are  one  to  several  orders  of  magnitude  greater  than  concentrations  in 
their  habitat  (Kinne,  1976).  As  external  NH3  concentrations  increase,  the  rate 
of  diffusion  outward  from  an  animal  decreases  and  toxicity  ensues  when  tolerable 
body  loads  are  exceeded.  Consequently,  the  toxicity  of  ammonia  to  aquatic  organ- 
isms is  generally  credited  to  the  NH3  molecule  (Ellis,  1937;  Wuhrmann  and 
Workers,  1948;  Downing  and  Merkens,  1955;  Spotte,  1970;  Hampson,  1976), 
despite  evidence  that  NH4+  adversely  affects  some  physiological  functions  (Shaw, 
1960;  Maetz,  1972;  Campbell,  1973).' 

The  chemistry  of  ammonia  in  solution  has  been  discussed  by  Wrhitfield  (1974) 
and  Colt  and  Tchobanoglous  (1976).  The  proportion  of  total  ammonia  existing  as 
NH3  is  dependent  on  temperature  and  ionic  strength  of  the  medium,  but  primarily 
on  the  pH  of  the  solution  (Warren,  1962;  Trussell,  1972;  Skarheim,  1973;  Whit- 
field,  1974;  Emerson,  Russo,  Lund  and  Thurston,  1975).  Calculations  by  these 
authors  show  that  the  NH3  fraction  of  ammonia  increases  as  pH  rises ;  an  increase 
of  one  pH  unit  elevates  the  NH3  concentration  tenfold.  As  previously  stated 
hypotheses  have  suggested,  the  toxicity  of  an  ammonia  solution  should  increase  at 
higher  pH  values. 

There  has  been  little  work  done  on  the  sensitivity  of  crustaceans  to  ammonia 
poisoning.  During  the  course  of  this  study  only  two  reports  were  found  which  give 

15 


16  ARMSTRONG,  CHIPPENDALE,  KXKillT  AND  COLT 

.systematic  evaluations  of  ammonia  toxicity  (based  on  mortality)  to  Crustacea 
(Anderson,  1944;  YYickins,  197(>),  and  a  single  study  investigating  the  sensitivity 
of  larval  crustaceans  to  ammonia  ( Delistraty,  Carlberg,  \  an  Olst  and  Ford,  1977). 
Adverse  effects  of  ambient  ammonia  on  some  physiological  functions  have  also  been 
reported  by  Shaw  (1960),  who  found  a  significant  reduction  in  sodium  influx  in 
the  crayfish  ^Istaciis  f>allipcs.  and  by  Mangum  ct  al.  (1976),  who  reported  reduced 
ammonia  excretion  rates  in  the  blue  crab,  Callinectes  sapidns.  The  exposure  levels 
of  ammonia  in  these  experiments  were  high,  18  and  180  mg  NH4+/liter,  respec- 
tively, and  may  have  been  approaching  lethal  concentrations.  However,  the 
authors'  interests  were  in  impairment  of  physiological  functions,  and  gross  signs 
of  stress  or  mortalities  at  these  concentrations  were  not  discussed. 

It  seems,  therefore,  that  most  researchers  investigating  the  effects  of  ammonia 
on  organisms  tend  to  concentrate  on  one  molecular  form  or  another  in  designing 
and  analyzing  their  work.  On  the  one  hand,  those  interested  in  concentrations 
lethal  to  fish  and  crustaceans  underline  the  importance  of  the  NHs  species  because 
of  its  ease  in  diffusing  across  membranes.  Consequently,  toxic  levels  of  up  to  a  few 
mg  NH3/liter  may  represent  well  over  100  mg  NH44/liter,  especially  at  pH  <  8.0. 
Such  high  ammonium  ion  concentrations  may  well  contribute  to  observed  mortality 
and  should  not  be  ignored. 

On  the  other  hand,  physiologists  concerned  with  the  interactions  between 
XH4+  in  salt  transport  processes  fail  to  address,  first,  the  possibility  that  the 
portion  of  the  high  total  ammonia  concentrations  used  (Carrier  and  Evans,  1976; 
Mangum  ct  al.,  1976;  Towle,  Palmer  and  Harris.  1976)  may  constitute  a  severe 
stress  to  an  organism  or  cellular  system,  thereby  affecting  a  process  that  is  thought 
to  be  NH4+-mediated  only ;  and  secondly,  the  effect  that  NH4+  inhibition  of  Na+ 
transport  or  ammonia  excretion  may  have  on  survival  of  organisms  in  different 
habitats. 

The  following  study  was  performed  to  determine :  first,  concentrations  of  am- 
monia lethal  to  larval  Macrobrachium  roscnbcrgii  in  short-term  exposures; 
secondly,  roles  and  interaction  of  NH.-j  and  NH4+  in  affecting  toxicity  using  pH  as 
a  variable,  and  to  analyze  any  observed  interaction  in  light  of  possible  physiological 
mechanisms;  and  thirdly,  sublethal  effects  during  short-term  exposure  using 
growth-reduction  as  the  criterion  of  toxicity.  An  additional  motive  underlying 
this  study  was  to  gain  information  on  ammonia  toxicity  that  could  be  applied  to 
general  water  quality  requirements  for  crustaceans.  This  is  particularly  important 
since  ambient  ammonia  concentrations  in  culture  or  holding  water  may  often  exceed 
levels  recommended  as  safe  (Spotte,  1970)  despite  extensive  filtration. 

MATERIALS  AND  METHODS 
Animals 

Larvae  were  produced  by  second  generation  U.  C.  Davis  brood  stock  initially- 
obtained  from  Hawaii  and  Thailand.  Broods  were  hatched  and  mass-reared  in  80 
liter  glass  aquaria  with  water  circulated  through  biological  filters.  Water  tempera- 
ture and  salinity  were  27-28°  C  and  \2%o  (Instant  Ocean  salts).  Larvae  were 
fed  newly  hatched  Arteuiia  salino  nauplii  and  were  used  in  tests  from  three  to  eight 
days  after  hatching. 


AMMONIA  TOXICITY  To  LARVAL  SI  Ik  I  Ml'  17 

Lcl/ial  to.vicity  biuussuys 

Static  bioassays  to  assess  ammonia  toxicity  were  performed  as  described  by 
Armstrong,  Stephenson  and  Knight  (1976a'»,  for  nitrite  toxicity  experiments  wtih 
Macrobracliiitni.  Fifteen  larvae  were  placed  in  each  250  beaker  containing  200  ml 
of  test  solution;  the  ratio  of  dry  weight  animal  biomass  (shrimp  +  Artemia)  to 
volume  of  solution  ranged  from  3  to  17.5  mg/liter.  Ammonia  concentrations  were 
made  by  serial  dilution  of  reagent  grade  XH4C1  (  Mallinckrodt)  for  a  concentration 
range  of  1.0-320  mg  ammonia/liter,  spaced  in  threefold  increments  per  decade. 
All  amonia  concentrations  and  controls  at  each  pH  were  replicated,  and  the  experi- 
ment was  run  twice  with  two  broods  of  larvae. 

Test  solutions  were  renewed  every  24  hr  at  which  time  larvae  were  transferred 
to  new  beakers,  mortalities  recorded,  and  fresh  brine  shrimp  added  to  give  a  density 
of  about  4-6  nauplii/ml.  After  the  initiation  of  an  experiment,  mortalities  were 
checked  at  30  min.  1,  2,  4,  8,  16,  and  24  hr.  and  three  times  in  each  subsequent  24 
hr  interval  to  the  conclusion  of  144  hr.  In  the  first  24  hr  death  was  defined  as  the 
cessation  of  heart  beat  and  pulsing  of  the  posterior  intestine.  Thereafter,  the 
opaqueness  commonly  developed  by  moribund  and  dead  larvae  was  used  as  the 
criterion  of  death  (Armstrong  ct  a!.,  1976a;  Armstrong,  Buchanan,  Mallon,  Cald- 
well,  and  Millemann,  1976b). 

Test  water  was  maintained  at  28°  C  (all  beakers  held  in  a  single  water  bath) 
and  12% c  salinity.  The  photoperiod  was  9D :  15L.  Three  pH  values  tested  were 
6.8,  7.6,  and  8.4.  Stock  water  of  \2%o  was  held  in  20  liter  carboys,  aerted  and  ad- 
justed frequently  to  desired  pH  with  1  M  NaOH  or  HC1  until  levels  stabilized 
which  required  several  days  prior  to  a  test.  The  pH  of  test  solutions  was  checked 
three  times/24  hr  period  and  adjusted  with  0.1  M  XaOH  or  HC1.  Solutions  to 
which  high  ammonia  concentrations  were  added  (>  100  mg  ammonia/liter)  re- 
quired pH  adjustment  immediately.  A  Corning  model  12  pH  meter  with  Ag/AgCl 
and  calomel  electrodes,  standardized  with  XBS  type  buffers,  was  used  for  mea- 
surements. Some  investigations  suggest  inaccuracies  in  measuring  pH  of  high 
ionic  strength  solutions  on  meters  calibrated  with  low  ionic  strength  buffers.  Hans- 
son  (1973)  stated  such  error  could  be  0.09  pH  units  at  20#r,  and  Whitfield  (cited 
in  Wickins,  1976)  reported  a  0.05  unit  error  at  35#c.  Since  our  salinity  was  I2r/(c, 
we  do  not  consider  the  possible  error  due  to  calibration  with  XBS  type  buffers  to 
be  significant.  The  mean  pH  values  (calculated  by  converting  pH  to  hydrogen  ion 
concentration,  averaging  and  then  returning  the  means  and  standard  deviation  to 
pH  units)  were  6.83  ±  0.09,  7.60  +  0.09,  8.34  ±  0.06  (n  >  100  for  each  pH). 
These  values  were  used  to  calculate  tin-ionized  ammonia  concentrations. 

Beakers  were  not  aerated ;  yet  dissolved  oxygen,  measured  with  a  Beckman  Oo 
Analyzer,  exceeded  90%  of  saturation  (7.3  mg/liter  at  temperature  and  salinity 
used)  after  24  hr  in  all  pH  and  ammonia  concentrations.  High  dissolved  oxygen 
values  were  due  to  the  change  of  solutions  every  24  hr,  low  biomass  to  volume  ratio, 
and  our  procedure  of  stirring  the  solution  of  beakers  several  times  a  day  to  check 
for  deaths. 

Xitrite  was  measured  in  representative  concentrations  from  all  three  pH  values 
by  a  sulfanilamide-based  colorimetric  reaction  (Federal  \Yater  Pollution  Control 
Administration,  1969).  At  the  end  of  24  hr,  the  average  nitrite  concentration  wa> 


IS  AUMSTKOXC,   C'HIPPENDALE,    KXKillT   AND   COLT 

9.4  ±  3.3  /zg  N(  );.-X/liter,  which  is  several  orders  of  magnitude  lower  than  the 
incipient  lethal  level  of  3  nig  X(  )--X/liter  reported  l>y  Armstrong  cl  <//.  (1976a)  for 
Macrobrachiunt. 

Ammonia  was  measured  with  an  (  )rion  Ammonia  Electrode  Model  95-10 
coupled  with  the  Corning  pH  meter.  Merks  (1975)  states  that  this  probe  loses 
accuracy  with  increasing  salinity,  and  correction  factors  must  be  used.  However, 
we  made  ammonia  standards  with  fresh  water  and  \2'/,f  sea  water  and  found  no 
difference  in  millivolt  readings  for  the  same  ammonia  concentrations  in  the  two 
media.  The  average  ammonia  concentation  at  24  hr  in  control  beakers  of  all  pi  I 
values  was  0.45  ±0.11  mg  ammonia/liter.  By  the  end  of  a  24  hr  period  the  change 
in  ammonia  levels  in  test  beakers  was  minimal.  Average  measured  concentrations 
were  102%  of  time-zero  nominal  levels,  indicating  little  volatilization  or  nitrifica- 
tion of  the  chemical  during  tests. 

Grozvth  experiments 

Larvae  of  this  warm  water  species  grow  rapidly,  molting  and  gaining  substantial 
weight  in  five  to  seven  days  (George,  1969;  Armstrong  ct  al.,  1976a).  Therefore, 
documentation  of  sublethal  effects  by  studying  growth  seemed  feasible  during  short- 
term  exposures.  After  establishing  lethal  concentrations  of  ammonia  at  each  pH, 
identical  bioassays  were  performed  using  two  sublethal  concentrations  per  pH.  A 
time-zero  sample  of  25  larvae  was  dried  at  70°  C  for  24  hr.  Animals  were  then 
individually  weighed  on  a  Cahn  Model  4700  automatic  electrobalance,  accurate  to  a 
few  p.g.  Test  animals  were  exposed  as  previously  outlined  and  at  the  conclusion  of 
a  test  were  dried  and  weighed  individually.  A  relative  growth  rate  wras  calculated 
for  each  treatment  with  the  formula  of  Waldbauer  (1968)  :  GR  --  P/TM,  where  P 
is  mean  dry  weight  gain  between  sampling  period,  T  is  time  between  sampling 
period  and  M  is  mean  individual  weight  over  the  sampling  period.  Two  sublethal 
growth  experiments  were  done :  the  first  with  five-day  old  larvae  exposed  to  treat- 
ments for  five  days;  and  the  second  with  three-day  old  animals  exposed  for  seven 
days. 

Statistical  analyses 

The  effect  of  ammonia  concentration  on  survival  was  investigated  using  a  three- 
way  analysis  of  variance.  The  effects  of  brood,  pH,  concentration  and  their  inter- 
actions on  the  dependent  variable,  time  to  death  for  each  larva,  were  analyzed. 
Effects  of  sublethal  concentrations  on  growth  were  investigated  with  one-way 
ANOVA,  by  treating  each  pH-concentration  combination  as  a  separate  factor.  If  a 
significant  F  value  (P  <  0.01  )  was  obtained,  treatment  differences  were  contrasted 
by  means  of  a  Q  value  (Snedecor  and  Cochran,  1967;  these  authors  regard  this 
contrasting  procedure  as  a  conservative  gauge  of  true  differences).  LC5o  values 
(the  concentration  of  toxicant  lethal  to  5Qc/r  of  the  test  organisms  in  a  specified 
time  period)  were  derived  from  log-probit  plots  of  concentration  vs.  mortality. 
LT50  values  (the  time  required  for  death  of  50%  of  the  organisms  in  a  given  con- 
centration of  toxicant)  were  obtained  by  probit  analysis  program  BMD/O3S 
(Dixon,  1970). 


AMMONIA  TOXICITY  TO  LARVAL   SHRIMP  19 

Calculation  of  un-ionised  ammonia:  NH$ 

The  NH3  fraction  of  the  total  ammonia  measured  is  calculated  from  the  general 
formula  for  bases  (Albert,  1973  J  : 

[Ammonia] 

!    _|_    10(pKa-PH) 

The  measurement,  and  possible  inaccuracies,  of  ammonia  and  pH  have  been  dis- 
cussed. The  pKa  remains  as  the  major  variable  of  the  equation  and  is  influenced 
by  physical  conditions  of  the  solution.  Emerson  ct  al.  (1975)  found  the  tempera- 
ture dependence  of  the  pKa  value  to  be  : 

pKa  =  0.09018  +  2729.92/T  (2) 

T  =  degrees  Kelvin 

Equation  (1)  is  based  on  an  infinite  dilution  model  for  which  the  activity  of  an  ion 
approaches  its  analytical  concentration  as  the  solute  concentration  approaches  zero. 
For  freshwater  systems  such  a  model  is  accurate.  However,  as  the  solute  concen- 
tration (i.e.,  salinity)  of  a  solution  increases,  the  activity  of  ions  and  uncharged 
species  may  be  significantly  different  from  their  concentration.  In  turn,  such 
changes  will  affect  pKa  values  and,  in  the  case  of  equation  (1),  will  consequently 
change  the  concentration  of  tin-ionized  ammonia  calculated. 

The  pKa  values  of  the  ammonia  system  in  sea  water  have  not  yet  been  experi- 
mentally determined.  Whitfield  (1974)  developed  theoretical  pKa  values  for  sea 
water,  but  did  not  calculate  them  for  salinities  less  than  about  20/£e.  The  salinity 
of  our  tests  was  \2%c  (ionic  strength,  I  --  0.242)  for  which  an  appropriate  pKa 
value  was  derived. 

The  acid  dissociation  reaction  for  ammonia  in  water  is  : 


NH4+  +  nH,O^  NH3-nH,0  +  H+  (3) 

The  equilibrium  expression  for  this  reaction  is  : 


|NH3-nH,0! 


{NH4+}{H80}« 
where  Ka  =  acidity  equilibrium  constant 

{i}   =  activity  of  the  ith  species 
Rewriting  equation  (4)  partially  in  terms  of  concentration 


„  j 

'  [NH4+]TNH4HH20}" 

where  {i}  =  yi[i] 

[i]  :=  concentration  of  the  ith  species. 

Since  the  electrode  method  for  determining  pi  I  measures  the  activity  of  the  hydro- 
gen ion  rather  than  concentration,  it  is  convenient  to  retain  the  {H*}  term.     Re- 


20  ARMSTRONG,  CHIPPENDALE,  KNIGHT  AND  COLT 

writing  equation  (5) 

Ka.{H2Oi"7NH/       [NH3]{H+{ 

TNH3  [NH4+] 

The    right    hand   expression   is    called    the    "mixed   acidity    equilibrium    constant" 
(Stumm  and  Morgan,  1970). 
Let 


K'a  -  (7) 

7NH:i 

Taking  the  logio  of  both  sides  and  making  the  substitution  that  pK  =  -log  K,  the 
following  equation  results  : 

pK'a  =  pKa  --  log7NH4+  +  log7XHa  -•  n  log  {H2O}  (8) 

The  values  used  for  the  right  hand  terms  are  as  follows:  pKa  =  9.154  (Emerson 
ct  al,  1975)  ;  --  log  yNH4+  =  0.140  (Stumm  and  Morgan,  1970;  Whitfield,  1974)  ; 
log  yNH3  =  0.008  (Whitfield,  1974);  -3  log  {H2O}  "  0.008  (Robinson,  1954). 
The  pK'a  calculated  for  28°  C  and  \2%0  was  9.310  and  was  used  in  equation  (1). 

Effect  of  NH3  and  NHf 

To  test  the  hypothesis  that  NH3  is  solely  responsible  for  ammonia  toxicity,  the 
concentration  of  total  ammonia  was  varied  with  pH  to  achieve  equal  levels  of  NHs 
but  unequal  levels  of  NH4+.  As  an  example,  using  equation  (1)  and  pK'a  =  9.31, 
it  is  calculated  that  10.3  mg  ammonia/liter  (9.3  mg  NH4+/liter)  will  give  1.0  mg 
NH3/liter  at  pH  8.34.  But  at  pH  6.83,  303  mg  ammonia/liter  (302  mg  NH4y 
liter)  is  required  for  the  same  concentration  of  NH;i.  Survival  was  monitored  to 
learn  if  these  widely  divergent  NH4+  concentrations  affected  larvae. 

RESULTS 

Analysis  of  variance  of  mortality  data  showed  no  significant  effect  due  to  brood 
or  any  interaction  involving  brood  and  therefore,  all  data  were  combined  for  com- 
putation of  LCso  and  LT->o  values.  There  was  a  highly  significant  effect  (P  < 
0.01)  due  to  both  pH  and  ammonia  concentration  and  the  interaction  of  these 
variables.  However,  during  the  four  bioassays  performed  (lethal  and  sublethal), 
survival  of  control  larvae  at  each  pH  always  exceeded  85%  and  averaged  95%  for 
144  to  168  hr  exposures. 

The  toxicity  of  ammonia  over  a  range  of  identical  concentrations  was  greatly 
influenced  by  the  pH  of  the  media.  The  24  hr  LCr,o  values  were  200,  115  and  37 
mg  ammonia/liter  at  pH  6.83,  7.60  and  8.34,  respectively  (Fig.  1),  and  the  sen- 
sitivity to  ammonia  remained  greatest  at  higher  pH  values  throughout  the  tests. 
By  144  hr  the  LCr.o  values  at  the  same  pH  values  had  decreased  to  80,  44,  and  14 
mg/liter;  approximately  a  2.7  fold  decrease  from  24  hr  values  (Fig.  1).  At  the 
test's  conclusion,  slopes  of  toxicity  curves  for  ammonia  in  solution  at  pH  8.34  and 
7.60  were  approaching  asymptotes  indicative  of  incipient  LC,-,o  values  (Sprague, 


AMMONIA  TOXICITY  TO  LARVAL   SHRIMP 


21 


300|- 


200 


<     100 

z 

o 


CP 

E 


50- 


20- 


10 


_pH 
6.83 

7.60 


8  34 


12     24 


48 


72  96 

TIME   (  MRS) 


120 


144 


FIGURE  1.  The  toxicity  of  total  ammonia  to  larval  -I/,  rosenbergii  exposed  in  solutions  of 
different  pH.  Bars  are  ±  one  standard  deviation. 

1969).  However,  no  such  decrease  in  the  slope  of  the  pH  6.83  toxicity  curve  had 
occurred,  indicating  longer  tests  were  needed  to  estimate  incipient  concentrations. 

The  time  to  death  for  larvae  held  in  solutions  of  equal  ammonia  concentration 
hut  different  pH  is  shown  in  Figure  2.  In  100  ing  ammonia/liter,  50%  mortality 
of  larvae  in  solutions  of  pH  8.34,  7.60,  and  6.83  occurred  in  about  9,  27,  and  125 
hr,  respectively.  Survival  of  larvae  held  in  32  mg/liter  at  pH  7.60  and  6.83  was 
nearly  equal  to  that  of  controls.  However,  animals  exposed  to  the  same  ammonia 
concentration  at  pH  8.34  were  all  dead  by  48  hr  (Fig.  2). 

Un-ionized  ammonia  was  not  the  exclusive  toxic  agent  in  these  tests,  and  the 
XH4+  molecule  apparently  contributed  to  mortality  also.  When  LCjo  values  were 
based  on  the  concentration  of  XH3  only  (I.e.,  normalized  with  respect  to  pH), 
there  was  no  equality  of  the  levels  found  to  be  toxic  at  specific  time  intervals  (Fig. 
3).  In  fact,  larvae  exposed  to  the  lowest  levels  of  XH3  (pH  =  6.83)  were  the  most 
susceptible  to  toxicity  due  to  the  concomitantly  high  levels  of  XH4+  present  (Table 
I).  The  proportions  of  XH3  and  XH4+  found  to  be  toxic  were  inversely  related  as 
pH  changed.  Consequently,  about  five  times  less  XH3  was  lethal  in  a  given  period 
at  low  pH  compared  to  the  high,  but  it  was  accompanied  by  six  times  more  XH4+ 
ion  (Table  I). 

The  effect  of  XH4(  on  survival  times  of  larvae  is  further  demonstrated  when 
the  response  of  groups  in  equal  XH3  concentrations  at  different  pH  values  was 
compared.  The  LT00  values  for  larvae  exposed  to  0.98  mg  XH3/liter  at  pH  6.83 


22 


ARMSTRONG,  CHIPPENDALE,  KNIGHT  AND  COLT 


98 


90 


§    50 


30 


10 


A 
• 
o 

A 


pH 

8.34 
7.60 
6.83 
8  34 
76,68 


CONCENTRATION 

100  mg/  I 
100  mg/l 
100  mg/  I 

32  mg/l 

32  mg/l 


8     10 


20 

TIME   (MRS) 


30     40 


60     80        120     160 


FIGURE  2.  Cumulative  percentage  of  mortality  of  M.  roscnbcrgii  larvae  exposed  to  several 
combinations  of  ammonia  and  pH.  Depicted  are  data  for  a  single  brood.  Survival  was  adjusted 
to  that  of  controls  which  averaged  95%  at  the  end  of  an  experiment. 


and  8.34  were  9  hr  and  >144  hr.  respectively,  while  the  corresponding  NHV  con- 
centrations were  319  and  9  mg/liter  (Fig.  4).  Animals  exposed  to  10.2  mg  NHa/ 
liter  survived  twice  as  long  as  those  exposed  to  5.5  mg/liter,  but  the  NH4+  con- 
centration was  3.5  times  higher  in  the  latter  case  (Fig.  4). 

Growth  of  Macrobrachium  larvae  was  reduced  in  sublethal  concentrations  of 
ammonia  and  also  seemed  to  be  influenced  by  levels  of  NH4+  rather  than  NH8. 
There  was  no  significant  effect  of  treatments  in  the  first  growth  experiment.  The 
initial  mean  weight  was  52  ±  5  /xg/larva  and  the  final  mean  weight  for  all  treat- 
ments was  77  ±  6  /tg/larva,  a  48^0  increase.  In  the  second  test,  with  smaller  larvae 
exposed  for  a  longer  period,  there  was  reduced  gowth  (P  <  0.01)  in  solutions  of 
32  mg  ammonia/liter  at  pH  6.83  and  7.60  (Table  II).  The  initial  weight  of  three- 
day  old  animals  was  35  ±  6  /xg  each.  At  the  test's  conclusion,  larvae  of  control 
groups  weighed  about  77  p.g  each  (a  120%  increase),  while  those  in  32  mg/liter 
averaged  55  and  61  /xg/larva  (57%  and  74%  increases)  in  pH  6.83  and  7.60,  re- 
spectively. These  weights  were  significantly  less  than  those  of  controls  (P  <  0.05, 

TABLE  I 

Concentrations  of  ammonia  toxic  to  M.  rosenbergii  larvae  expressed  as   both  the  NH3  and 
molecules.* 


24  hr  LCso  (mg/liter) 

144  hr  LCto  (mg/liter) 

oH 

NH8 

NH4+ 

NH3 

NH<  + 

6.83 

0.66 

199.34 

0.26 

79.74 

7.60 

2.10 

112.90 

0.80 

43.20 

8.34 

3.58 

33.42 

1.35 

12.65 

*  Total  ammonia  =  [NH:j]  +  [NH4+]  and  is  depicted  in  Figure  1. 


AMMONIA  TOXICITY  TO  LARVAL   SHRIMP 


23 


Q  statistic)  and  were  the  only  important  differences  found.  Reduction  in  growth 
was  not  correlated  with  XH3  concentrations.  The  relative  growth  rate  (GR)  of 
controls  was  0.108  g/(g  dry  body  wt-day)  (Table  II).  Larvae  exposed  to  the 
highest  XH3  concentration  of  0.98  mg  Nils/liter  (XH4+  ==  9  mg/liter)  had  a  GR 
=  0.097,  while  those  exposed  to  0.11  mg  XHs/liter  (XH4+  -  31.9  mg/liter)  had 
a  GR==  0.063  (Table  II). 

DISCUSSION 

The  toxicity  of  ammonia  to  Macrobrachium  larvae  is  inextricably  linked  to  the 
pH  of  a  solution,  the  total  ammonia  concentration  present,  and  the  proportions  of 
that  total  which  exist  as  either  XH3  or  XH4+.  Undoubtedly  other  factors,  such 
as  dissolved  oxygen  and  salinity,  could  be  varied  from  optimal  levels  to  further  com- 
plicate the  story  of  ammonia  toxicity  to  this  crustacean. 


CP 

E 


0.3- 


02 


12 


48 


72  96 

TIME   (HRS; 


120 


144 


FIGURE  3.  Concentrations  of  uii-iunized  ammonia  (NHa)  causing  50%  mortality  in  various 
time  intervals.  The  NH3  concentrations  account  for  about  10%,  2%,  and  0.3%  of  the  total 
ammonia  levels  in  the  high  to  low  pH  values,  respectively. 


24 


ARMSTROXC,   rHlPPKXDALK,   KXKiHT  AND  COLT 


140 

120 

100 
80 

60 

40 
30 

20 

C/5 

tr       16 

x 

2      12 


0 


2.9 


997 


CONCLUSION    OF    TESTS 

PH 
E3    8.4 

["I    6.8 


319  mg  NH4/I 


90 


314  5 


0.31 


098 


5.5 


102 


NH, (mg/ 


FIGURE  4.  Time  to  50%  mortality  of  larvae  exposed  to  several  concentrations  of  NH3 
ammonia  at  different  pH.  At  the  top  of  bars  are  concentrations  of  NH.i+.  Bars  exceeding  144 
hr  had  survival  equal  to  controls  by  the  end  of  the  experiment. 

As  is  traditionally  done  in  fish  bioassays  with  ammonia,  toxic  concentrations 
derived  from  the  present  tests  could  not  be  normalized  to  pH  variations  by  ex- 
pressing results  in  terms  of  the  NH3  molecule  only,  because  the  NH4+  ion  figured 
critically  in  causing  stress.  Toxic  ammonia  concentrations  differ  between  the  high 
and  low  end  of  the  pH  range  tested,  and  are,  we  believe,  determined  by  NH3  at 
high  pH  and  NH4+  at  low  pH  values.  At  a  pH  of  8.34  the  incipient  LC50  value 
was  estimated  to  be  14  mg  ammonia/liter,  which  are  NH:{  and  NH4+  proportions  of 
1.35  and  12.65  mg/liter,  respectively  (Table  I).  Growth  at  this  same  pH  was  not 
inhibited  by  10  mg  ammonia/liter,  indicating  that  an  incipient  lethal  level  is  indeed 
about  12-14  mg/liter.  In  solutions  of  lower  pH,  more  total  ammonia  is  required 


AMMONIA  TOXICITY  TO  LARVAL  SHRIMP 


25 


to  cause  toxicity,  and  the  NHa  fraction  of  these  concentrations  decreases  exponenti- 
ally with  pH.  Using  growth  as  a  sensitive  gauge  of  stress,  32  mg  ammonia/liter 
retarded  development  at  both  pH  6.83  and  7.60.  The  un-ionized  NH:{  fraction  at 
pH  6.83  is  0.11  mg  NH^/liter,  only  0.3%  of  the  total  concentration  and  about  11 
times  less  than  the  incipient  LC.-,o  value  derived  for  pH  8.34.  XH4"  ion  accounts 
for  nearly  all  ammonia  present  and  is  the  species  of  ammonia  probably  responsible 
for  toxicity  at  low  pH  values. 

These  observations  may  be  combined  in  a  model  (Fig.  5)  to  describe  differ- 
ential ammonia  toxicity  caused  by  changes  in  pH.  Water  conditions  shown  in  the 
model  are  those  actuallv  measured  in  these  tests.  Values  for  chemical  factors  in 

j 

larval  blood  have  been  assumed  based  on  literature  data  for  adult  and  juvenile 
crustaceans.  Blood  osmolarity  was  estimated  to  be  500  mOsmol  ==  15. 8f/cc  salinity 
based  on  determinations  made  with  M.  rosenbcrgii  post-larvae  (Armstrong  and 
Nelson,  unpublished  data;  Sandifer,  Hopkins  and  Smith,  1975).  Blood  pH  was 
chosen  to  be  7.55,  7.65,  and  7.75  at  corresponding  water  pH  values  of  6.83,  7.60, 
and  8.34  (from  data  of  Johansen,  Lenfant  and  Mecklenburg,  1970;  Truchot,  1975; 
Weiland  and  Mangum,  1975;  Mangum  ct  al.,  1976).  Total  blood  ammonia  was 
taken  to  be  representative  of  levels  in  control  larvae,  treated  as  described,  before 
addition  of  toxic  concentrations  of  ammonia  to  the  ambient  water.  A  concentra- 
tion of  12  mg  ammonia/liter  of  blood  was  assumed  from  data  of  Myers  (1920), 
Florkin  and  Renwart  (1939),  Florkin  and '  Frappez  (1940),  Gifford  (1968),  and 
Mangum  ct  al.  (1976).  Sodium  influx  is  depicted  as  relative  magnitudes  varying 
with  ambient  NH4+  concentrations.  The  pKa,  9.33,  used  to  calculate  un-ionized 
ammonia  in  the  blood,  was  detemined  for  a  salinity  of  16/^r,  as  previously  outlined. 
The  model  (Fig.  5)  proposes  that  larvae  exposed  to  ammonia  at  higher  pH 
(«  8.4)  will  be  most  affected  by  NH:i,  which  is  nonpolar  and  can  readily  diffuse 
through  biological  membranes  such  as  the  gills  (Warren,  1962).  Of  the  total  am- 

TABLE  1 1 
Effect  of  ammonia  on  the  relative  growth  rate  of  Macrobrachiuni  larvae  held  in  water  of  different  pH. 


pH 

Total  ammonia 
XH3  +  XH4+ 

(mg/liter) 

Un-ionized  ammonia 
NH, 
(mg/liter) 

GR* 
g/(g  body  wt-day) 

Final  mean** 
dry  weight 
(±s.d.) 
/ig  /larva 

6.83 

0 

0 

0.107 

77  (15) 

10 

0.03 

0.105 

76  (11) 

32 

0.11 

0.063*** 

55  (11) 

7.60 

0 

0 

0.107 

77  (16) 

10 

0.20 

0.113 

81  (15) 

32 

0.63 

0.077*** 

61  (13) 

8.34 

0 

0 

0.109 

78  (11) 

3.2 

0.31 

0.091 

68  (12) 

10 

0.98 

0.097 

71  (14) 

*  GR  =  P/TM  for  dry  wt.  See  text  for  explanation. 

**  Seven  day  exposure;  initial  mean  dry  weight  =  35  ±  6  /ig 'larva ;  n  =  19-23  larvae  per 
group. 

***  Significantly  different  from  controls,  P  <  0.05. 


26 


ARMSTRONG,  CHIPPENDALE,  KNIGHT  AND  COLT 


SALINITY 


PH 


TOTAL  AMMONIA 
(mg/l) 


[NHjmg/l 
(DIFFUSION) 


[  NH^Jmg/l 

(COUNTER -ION 
ACTIVE  TRANS- 
PORT) 


No    INFLUX 

(ACTIVE 

TRANSPORT) 


BLOOD 


WATER  BLOOD  WATER 


BLOOD 


500MOSM 


=  !5.8%c 


755 


12 


.20- 


II  8 


I2%0  = 

3IOOmgNa/l 


6.83 


81 


•.27 


80.7 


INHIBITS 


I5.8%o 


7.65 


12 


.25 


11.7 


12  % 


7.60 


40 


392 


1 


I  E     Q     O/ 

I  D .  O    /o  o 


7.75 


12 


WATER 


1 2  %, 


8  34 


14 


.31 


I  1.7 


12.6 


FIGURE  5.  A  proposed  mechanism  explaining  the  differential  effects  of  NH3  and  NH4+  on 
A/,  rosenbergii  larvae  cultured  in  solutions  of  different  pH.  The  values  for  blood  salinity,  pH, 
and  total  ammonia  were  estimated  from  literature  data  as  described  in  the  text.  These  condi- 
tions are  assumed  to  be  typical  of  larvae  prior  to  addition  of  high  ambient  ammonia.  The  water 
ammonia  levels  are  incipient  lethal  concentrations  derived  for  the  three  pH  values  tested.  Am- 
monia in  water  of  high  pH  exists  in  relatively  large  quantities  as  unionized  NH3,  which  rapidly 
diffuses  into  larvae,  increasing  blood  ammonia  to  toxic  levels.  In  low  pH  solutions  ammonia 
exists  almost  totally  as  NH4+.  This  ion  is  shown  to  compete  with  Na  in  active  transport  pro- 
cesses and  toxicity  ensues  from  osmoregulatory  failure. 

monia  found  toxic  at  high  pH  about  1.35  mg/liter  or  10%  exists  as  NHs.  This 
level  exceeds  that  postulated  for  the  blood  by  about  four-fold,  and  consequently  NH:i 
would  diffuse  into  animals.  At  a  blood  pH  "  7.65  the  molecule  would  be  pro- 
tonated  to  NH4+,  thereby  maintaining  the  NHS  diffusion  gradient  inward.  Body 
concentrations  of  ammonia  would  rise  if  alternate  routes  of  excretion  could  not 
expel  this  surplus,  and  toxicity  follow,  perhaps  via  a  mode  described  by  Campbell 
(1973).  Toxicity  might  include  elevation  of  blood  pH  as  NH3  is  protonated  and 
a  decrease  in  substrate  for  the  tricarboxylic  acid  cycle  as  excess  ammonia  reverses 
the  usual  oxidation  of  glutamate  (Campbell,  1973).  Toxicity  clue  to  inward  dif- 
fusion of  NH.s  at  high  pH  is  rapid  and  caused  mortality  among  test  larvae  in  2-18 
hr(Fig.2). 

The  deleterious  effect  of  high  ambient  ammonia  levels  on  an  alternate  route  of 
ammonia  excretion  from  the  blood  (nondiffusion)  is  the  second  component  of  the 
model.  It  is  proposed  that  inhibition  of  sodium  influx  is  a  major  factor  contribut- 
ing to  ammonia  toxicity  at  low  pH.  Larvae  in  water  of  pH  6.83  died  in  81  mg 


AMMONIA   TOXiriTY   TO   LARVAL   SHRIMP 

ammonia/liter,  which  is  a  XI  I-  concentration  of  only  0.27  ing/liter.  This  water 
concentration  is  nearly  equal  to  the  blood  level  estimated  and,  even  though  the  rate 
of  diffusion  of  NH3  outward  is  probably  reduced,  the  decrease  is  apparently  not 
serious.  [Recall  that  0.98  nig  XH3/liter  at  pH  8.34  caused  no  mortality  (Fig.  4) 
or  growth  inhibition  (Table  II),  yet  this  concentration  certainly  exceeded  blood 
levels  and  should  have  established  an  XH;i  diffusion  gradient  inward.]  Nearly  all 
of  the  ammonia  (80.7  mg/liter)  exists  as  the  XH4+  ion.  By  successfully  competing 
with  sodium  ions,  the  XH4+  would  both  reduce  the  influx  of  XV,  thereby  diminish- 
ing body  concentrations  of  this  important  salt,  and  also  cause  body  levels  of  am- 
monia to  rise  by  itself,  riding  the  transport  mechanism  in  or  preventing  metabolic 
XH4+  from  riding  it  out.  The  resistance  of  the  larvae  to  this  form  of  osmoregula- 
tory  inhibition  by  XH4+  is  apparently  greater,  and  toxic  manifestations  do  not 
develop  as  rapidly  as  when  copious  XHs  diffusion  inward  (high  pH)  is  operative. 
Mortality  occurred  in  40-140  hr  at  pH  6.83  (Fig.  2),  and  growth  inhibition  prob- 
ably requires  exposures  of  5-7  days  to  be  measurable  with  the  larval  stages  used. 

The  hypothesis  that  toxicity  at  low  pH  is  caused  by  inhibition  of  XV  transport 
by  XH4+  (Fig.  5)  has  been  based  on  several  studies.  Ammonium  ion  has  long 
been  suggested  as  a  counter-ion  for  Xa+  transport  (Krogh,  1939).  Recently 
Mangum  and  Towle  (1977)  discussed  the  physiological  roles  of  internal  XH4+  in 
the  euryhaline  blue  crab.  They  believe  XTH4+  aids  in  activating  gill  ATPase,  serves 
as  one  counter-ion  for  sodium  transport,  aids  in  maintaining  charge  balance  as  it  is 
excreted,  and  is  an  important  form  of  ammonia  in  which  this  toxicant  is  elimi- 
nated from  the  body.  In  the  external  milieu,  XTH4+  can  substantially  reduce  the 
influx  of  Xa+.  Shaw  (1960)  found  that  18  mg  XH4+/liter  caused  an  SOfi  de- 
crease in  XTa+  influx  rates  in  the  crayfish,  Astacus  pallipes.  Inhibition  of  Xa  trans- 
port across  gills  by  external  XH4+  and  stimulation  of  XV  uptake  after  intraperi- 
toneal  injection  of  XTH4+  has  also  been  documented  for  fish  (Maetz  and  Garcia- 
Romeu,  1964;  Carrier  and  Evans,  1976). 

An  interesting  aspect  of  the  XV-XH4+  transport  system  regards  the  affinity  of 
the  carrier  mechanism  for  either  molecule.  Shaw  (1960)  found  that  the  inhibition 
of  sodium  influx  caused  by  ambient  ammonium  ion  could  be  countered  by  increasing 
ambient  sodium  levels.  \Yorking  with  a  freshwater  crustacean  in  low  levels  of  both 
XV  and  XH4+,  Shaw  concluded  that  a  concentration  ratio  of  10:1  favoring  X"H4+ 
must  exist  for  inhibition  of  sodium  transport  to  occur,  and  that  the  affinity  of 
sodium  for  the  transport  site  is  greater  than  that  of  ammonium  ion.  The  present 
experiments  were  done  in  \2c/fc  sea  water  or  about  3100  mg  XV/h'ter  (Instant 
Ocean  salt  is  25.8%  X"a  by  weight  based  on  manufacturer's  analysis).  Based  on 
the  concentrations  of  XH4+  found  toxic  (32-80  mg  XTH4+/liter),  the  XH4+  to  XV 
ratios  in  our  tests  were  0.01-0.02:  1.  Such  low  ratios  for  XH4+  imply  that  the  ion 
has  a  greater  affinity  for  the  transport  site  than  XV,  contrary  to  Shaw's  conclusion. 
This  discrepancy  might  be  partially  explained  by  lower  affinity  of  the  transport 
mechanism  for  Xa+  in  the  euryhaline  Macrobrachiuin  than  in  the  freshwater  crayfish 
of  Shaw's  experiments.  The  Km  values  for  sodium  transport  may  be  tenfold 
greater  in  saline  species  than  in  similar  freshwater  forms  (Prosser,  1973).  Alter- 
natively the  low  XH4+ :  Xa+  ratios  may  indicate  that  XH4+  is  causing  toxicity  in  a 
manner  other  than  inhibition  of  sodium  movement. 


28  AKMSTKOXd.   UIII'I'KXDAUC.    KXKillT   AXI)   COM 

It  has  IK-CM  demonstrated  in  these  studies  that  sufficiently  high  concentrations 
of  NH4+  in  water  of  lo\v  pi  1  is  lethal  to  crustacean  larvae,  even  though  the  NH3 
concentration  present  may  he  sublethal.  A  model  ascrihes  such  toxicity  to  com- 
petitive inhibition  of  Xa+  transport.  It  is  probably  an  over-simplification  to  at- 
tribute the  toxicity  of  ammonia  only  to  NH;{  at  high  pH  and  to  NH4+  at  low  pH. 
There  may  he  a  contribution  from  each  species  at  a  total  ammonia  concentration 
found  to  be  toxic,  but  we  believe  our  model  is  accurate  in  assigning  the  bulk  of 
toxicity  to  either  NH3  or  NH4+  as  the  change  in  pH  influences  the  ratios  between 
them.  Accordingly,  we  offer  a  caution  for  those  studying  ammonia-induced  re- 
sponses in  organisms  to  consider  the  contribution  from  both  NHs  and  NH4+  species 
in  interpreting  results.  Relatively  low  but  lethal  concentrations  of  XH3  may  be 
accompanied  by  large  amounts  of  NH4+,  especially  at  lower  pH  values.  High  total 
ammonia  levels  used  in  some  physiological  experiments  may  represent  near-lethal 
concentrations  of  NH3,  particularly  at  higher  pH  values.  Mangum  ct  al.  (1976) 
reported  that  10  HIM  - :  180  mg  ammonia/liter  was  used  in  tests  on  ammonia  ex- 
cretion. At  a  pH  of  about  7.8,  this  would  equal  5.4  mg  NH,3/liter,  which  is  well 
within  the  range  we  found  to  be  toxic  (Table  I). 

Finally,  some  discussion  of  the  results  relative  to  water  quality  requirements  of 
crustaceans  is  warranted.  Whether  the  maintenance  of  animals  is  for  long  periods 
in  commercial  operations  or  for  short  acclimations  prior  to  physiological  experi- 
ments, water  quality  is  an  important  variable  that  should  be  monitored  and  regu- 
lated. Ammonia  concentrations  found  to  be  toxic  in  this  study  are  in  accord  with 
other  values  reported  at  similar  pH  levels.  Wickins  (1976)  found  that  101  mg 
ammonia/liter  (pH  --  7.0)  gave  an  LT50  of  24  hr  for  adult  Macrobrachium. 
Further,  growth  was  reduced  30-35^  in  concentrations  of  0.19-0.39  mg  NH3/liter, 
which  corresponds  to  a  very  high  range  of  20-41  mg  NH4+/liter  (pH  =  7.2,  pKa 
=  9.22  at  his  test  conditions).  Following  from  the  results  of  the  present  study,  we 
suggest  that  inhibition  of  growth  resulted  primarily  from  the  NH4  ion  and  not 
NH3,  as  reported  (Wickins,  1976).  Anderson  (1944)  reported  that  Daphnia 
magna  was  immo.bilized  in  16-24  hr  when  exposed  to  46  mg  ammonia/liter  (no  pH 
given)  ;  and  an  incipient  LC-,o  for  larvae  of  the  lobster,  Houianis  aniericanns,  was 
37  mg  ammonia/liter  at  pH  :=  8.1,  salinity  =  33.4#o  (Delistraty  et  al.,  1977).  The 
incipient  LC50  calculated  for  Macrobrachium  larvae  in  water  of  pH  7.60  was  40  mg 
amomnia/liter. 

These  toxic  concentrations  are  rather  high  and  greatly  exceed  the  "safe"  level 
of  0.1  mg  ammonia/liter  recommended  by  Spotte  (1970).  Larvae  in  the  present 
test  survived  10  and  <  32  mg  ammonia/liter  for  seven  days  at  pH  =  8.34  and  6.83, 
respectively.  Such  levels  would  probably  be  injurious  over  long  periods  and  an 
application  factor,  applied  to  the  incipient  LC.-,o  values  or  concentrations  inhibiting 
growth,  would  be  needed  to  estimate  safe  levels.  Sprague  (1971)  summarizes 
thought  on  this  topic  with  the  conclusion  that  0.1-0.3  of  an  incipient  LC.-)0  value  can 
predict  safe  concentrations.  Such  a  criterion  would  predict  as  safe  about  1  mg 
ammonia/liter  at  pH  8.34  and  3.2  mg/liter  at  the  lower  pH.  However,  the  lack 
of  mortality  and  sublethal  growth  inhibition  at  10  mg/liter  leads  us  to  conclude 
that  short-term  exposure  to  rather  high  ammonia  levels  may  not  be  damaging  to 
Macrobrachium. 

In  general,  the  use  of  flow-through  culture  systems  with  water  exchange  acle- 


AMMONIA  TOX1CITY  TO   LARVAL   SliRIMl'  29 

(juate  to  dilute  excreted  ammonia,  or  closed-systems  \vitli  conditioned,  nitrifying 
filters  for  detoxification  sin  mid  minimize  the  threat  «f  ammonia  toxicity  for  crusta- 
ceans. In  our  research  culture  facilities,  the  ammonia  concentration  in  water 
passed  through  biological  filters  averages  0.5  mg/liter  (pH^S.l),  well  below 
toxic  levels  reported  in  this  study. 


We  greatly  appreciate  the  critical  review  and  criticism  of  the  manuscript  given 
by  Drs.  J.  Crowe,  S.  Nelson  and  C.  Siegfried.  Dr.  P.  "Wilde  discussed  the  section 
on  water  chemistry  with  us,  and  L.  Shaw  gave  patient  help  with  statistical  analyses. 
This  research  was  supported  by  a  grant  from  the  State  of  California  to  the  Uni- 
versity of  California.  Davis,  Aquaculture  Group. 

SUMMARY 

1.  The  toxicity  of  ammonia  to  Macrobrachmm  larvae  was  tested  at  pH  6.83. 
7.60,  and  8.34,  and  the  respective   144  hr  LC5o  values  were  80,  44,  and    14  ing 
ammonia/liter. 

2.  Toxicity  of  ammonia  was  not  due  solely  to  the  XHs  molecule.     In  solutions 
of  different  pH  and  equal   XH:{  concentrations,   survival   was  greatly  reduced  as 
NH4+  levels  increased. 

3.  A  model  is  proposed  to  explain  the  differential  effect  of  ammonia  as  pH 
varies.     At  higher  pH   ( 8.4 )   toxicity  results  from  copious  diffusion  of  NHs  into 
larvae.    At  lower  pH  (6.8)  toxicity  is  thought  to  result  from  competitive  inhibition 
of  Na+  transport  by  NH4+. 

4.  Retardation  of  growth  was  documented  in  sublethal  concentrations  of  am- 
monia at  6.8  and  7.6.     The  average  dry  weight  was  about  26%   less  than  that  of 
controls  (P  <  0.05)  after  a  seven  day  exposure. 

5.  Results  are  discussed  relevant  to  the  culture  and  maintenance  of  crustaceans, 
and  it  is  concluded  that  ammonia  will  not  pose  a  substantial  threat  in  adequately 
managed  systems. 

LITERATURE  CITED 

ALBERT,  A.,  1973.     Selective  to.ricity.    Chapman  and  Hall,  London,  597  pp. 

ANDERSON,    B.    G.,    1944.      The   toxicity   thresholds   of   various    substances    found    in    industrial 

waste  as  determined  by  the  use  of  Daphnia  inac/iia.    Scicage  Works  /..  16:   1156-1165. 
ARMSTRONG,  D.  A..  M.  J.  STEPHENSON,  AND  A.  YV.  KNIGHT,  1976a.    Acute  toxicity  of  nitrite  to 

larvae  of  the  giant   Malaysian   prawn,   Macrobrachium   rosenbcrgii.     Aquaculture,   9: 

39-46. 
ARMSTRONG,  D.  A.,  D.  V.  BUCHANAN,  M.  H.  MALLON,  R.  S.  CALDWELL,  AND  R.  E.  MILLE- 

MANN,  1976b.    Toxicity  of  the  insecticide  methoxychlor  to  the  Dungeness  crab  Cancer 

iiKif/istcr.   Mar.  BioL,  38:  239-252. 

CAMPBELL,  J.  YV..   1973.     Nitrogen  excretion.    Pages  279-316  in  C.  L.  Prosser,  Ed.,  Compara- 
tive ani)iial  physiolm/y.     YV.  B.  Saunders  Co.,  Philadelphia. 
CARRIER,  J.  C.,  AND  D.  H.  EVANS,  1976.    The  role  of  environmental  calcium  freshwater  survival 

of  the  marine  teleost,  Lagondon  rhomhoidcs.    J.  Exp.  BioL,  65:  529-538. 
COLT,  J.,  AND  G.  TCHOBANOGLOUS,  1976.     Evaluation  of  the  short-term  toxicity  of  nitrogenous 

compounds  to  channel  catfish,  Ictalurus  f>unciatns.     Aquaculture,  8:  209-224. 


M)  ARMSTRONG.  CHIPPENDALE,   KNIGHT  AND  COLT 

DKI.I:VIUATV.  1).  A.,  J.  M.  CARLKKRG,  J.  C.  VAN  OLST,  AND  R.  F.  FORD,  1977.     Ammonia  toxicity 

in  cultured  larvae  of  the  American  lobster,  llomarus  americunus.     I'roe.  li'orld  MiiricuL 

Soc.,  8th  Annual  Meeting,  San  Jose,  Costa  Rica. 
DIXON,   W.   J.    (Ed.),    1970.     HMD   biomedical   computer   programs.      University   of   California 

Press,  Los  Angeles,  773  pp. 
DOWNING,  K.  M.,  AND  J.  C.  MERKENS,  1955.     The  influence  of  dissolved-oxygen  on  the  toxicity 

of  un-ionized  ammonia  to  rainbow  trout    (Salmo  gairdnerii  Richardson).     Ann.  Appl. 

Biol.,  43 :  243-246. 
KI.LIS,  M.  M..   1937.     Detection  and  measurement  of  stream  pollution.    Bull.   U.  S.  Bur.  Fish., 

48(22)  :  365. 

EMERSON,  K.,  R.  C.   Rrsso,  R.  LUND,  AND  R.  V.  THURSTON,   1975.     Aqueous  ammonia  equi- 
librium calculations:    effects  of  pH   and  temperature.     /.   Fish.   Res.   Board  Can.,  32: 

2379-2383. 
FEDERAL   WATER    POLLUTION    CONTROL   ADMINISTRATION,    1969.      Pages    181-187    in    FWPCA 

methods  for  chemical  analysis  of  i^atcr  u<asfc.     U.  S.  Dept.  Interior,  Washington,  D.  C. 
FLORKIN,  M..  AND  G.  FRAPPEZ,  1940.     Concentration  de  1'ammoniaque  in  vivo  et  in  vitro,  dans 

le  milieu  interieur  des  invertebres.     III.  Ecrevisse,   Hydrophile.    Dytique.     Arch.  Int. 

Physiol.,  50:  197-202. 
FLORKIN,  M..  AND  H.  RENWART,  1939.     Concentration  de  ammoniaque,  in  vivo  et  in  vitro,  dans 

le  milieu  interieur  des  invertebres.     II.     Escargot  et  Homard.    Arch.  Int.  Physiol.,  49: 

127-128. 
FROMM,  P.  O.,  AND  J.  R.  GILLETTE,  1968.     Effect  of  ambient  ammonia  on  blood  ammonia  and 

nitrogen   excretion   of   rainbow   trout    (Salmo   gairdnerii).      Comp.    Biochcin.    Physiol., 

26 :  887-896. 
GEORGE,  M.  J.,  1969.     Genus  Macrobrachium  Bate  1868.     Bull.  Cent.  Mar.  Fish.  Res.  Inst.,  14: 

179-216. 
GIFFORD,  C.  A.,  1968.     Accumulation  of  uric  acid  in  the  land  crab,  Cardisoma  guanhumi.    Am. 

Zoo/.,  8:  521-528. 

HAMPSON,  B.  L.,  1976.     Ammonia  concentration  in  relation  to  ammonia  toxicity  during  a  rain- 
bow  trout   rearing   experiment   in   a   closed   freshwater-seawater   system.    Aquaculture, 

9:  61-70. 
HANSSON,  I.,  1973.    A  new  set  of  pH-scales  and  standard  buffers  for  sea  water.    Deep  Sea  Res., 

20:  479-491. 
HARTENSTEIN,  R.,   1970.     Nitrogen  metabolism   in  non-insect  arthropods.     Pages  299-372  in  J. 

W.  Campbell,  Ed.,  Comparative  biochemistry  of  nitrogen   metabolism.    I.   The  inverte- 
brates. Academic  Press,  New  York. 
HOCHACKHA,   P.  W.,  AND  G.   N.   SOMERO,   1973.    Strategies  of  biochemical  adaptation.  W.   B. 

Saunders  Co.,  Philadelphia,  358  pp. 
JOHANSEN,  K.,  C.  LENFANT,  AND  T.  A.  MECKLENBURG,  1970.     Respiration  in  the  crab,  Cancer 

magistcr.  Z.  Vcryl.  Physio/..  70:  1-19. 
KIXNE,  O.,  1976.     Cultivation  of  marine  organisms:    water  quality  management  and  technology. 

Pages  79-300  in  O.  Kinne,  Ed.,  Marine  ecology,  J'ol.  Ill,  Part  1.    Wiley-Interscience, 

New  York. 
KROGH,    A.,    1939.      Osmotic    regulation    in    aquatic    animals.      Cambridge    University    Press, 

London. 
MAETZ,  J.,  1972.     Interaction  of  salt  and  ammonia  transport  in  aquatic  organisms.    Pages   105- 

154  in  J.  W.  Campbell  and  L.  Goldstein,  Eds.,  Nitrogen   metabolism  and  the  environ- 
ment. Academic  Press,  New  York. 
MAETZ,  J.,  AND  F.  GARCIA-ROMEU,  1964.     The  mechanism  of  sodium  and  chloride  uptake  by  the 

gills  of  a  fresh-water  fish,  Carassius  auratus.     II.     Evidence  for  NH4+/Na+  and  HCO3/ 

Cl-  exchange.    /.  Gen.  Physiol.,  47  :  1209-1227. 
MANGUM,   C.,  AND  D.   TOWLE,   1977.     Physiological  adaptation  to  unstable  environments.   Am. 

Sci.,  65 :  67-75. 
MANGUM,  C.  P.,  S.  U.  SILVERTHORN,  J.  L.  HARRIS,  D.  W.  TOWLE,  AND  A.  R.  KRALL,   1976. 

The  relationship  between  blood  pH,  ammonia  excretion  and  adaptation  to  low  salinity 

in  the  blue  crab,  Callinectes  sapidns.    J.  E.rp.  Zoo!.,  195:   129-136. 


AMMONIA  TOXICITY  TO  LARVAL  SHRIMP  31 

MKRKS,  A.  G.,   1975.     Determination  of  ammonia  in  sea  water  with  an  ion-selective  electrode. 

Neth.  J.  Sea  Res.,  9 :  371-375. 
MYERS,  R.  G.,   1920.     A  chemical  study  of  the  blood  of  several  invertebrate  animals.    /.  Biol. 

C 'hem.,  41:  119-147. 
PROSSER,   C.   L.,    1973.     Water :  osmotic  balance :  hormonal   regulation.      Pages   1-78   in   C.   L. 

Prosser,  Ed.,  Comparative  animal  physiology.    W.  B.  Saunders  Co.,  Philadelphia. 
ROBINSON,   R.  A.,   1954.     The  vapor  pressure  and  osmotic  equivalence  of   seawater.     /.   Mar. 

Biol.  Assoc.  U.K.,  33  :  449-455. 
SANDIFER,  P.  A.,  J.  S.  HOPKINS,  AND  T.  SMITH,  1975.     Observations  on  salinity  tolerance  and 

osmoregulation  in  laboratory-reared  Maerobraehiitm  roscnbcrgii  post-larvae  (Crustacea: 

Caridea).    Aquaculture,  6:  103-114. 
SHAW,  J.,   1960.     The  absorption  of  sodium  ions  by  the  crayfish  Astacns  pallipcs  Lereboullet. 

III.  The  effect  of  other  cations  in  the  external  solution.   /.  E.vp.  Biol..  37:  548-556. 
SKARHEIM,  H.  P.,  1973.     Tables  of  the  fraction  of  ammonia  in  the  undissociated  form  for  pH 

6-9,  temperature  0-30°  C,  TDS  0-3000  mg/ liter,  and  salinity  5-35  g/kg.  SERL  Report 

Number  73-5,  University  of  California,  Berkeley. 
SNEDECOR,  G.  W.,  AND  W.  G.  COCHRAN,  1967.    Pages  272-273  in  Statistical  methods.  The  Iowa 

State  University  Press,  Ames,  Iowa. 
SPOTTE,    S.    H.,    1970.     Fish   and  invertebrate   culture,    water   management    in    closed   systems. 

Wiley-Interscience,  New  York,  New  York,  145  pp. 
SPRAGUE,  J.  B.,   1969.     Measurement  of  pollutant  toxicity  to   fish.     I.   Bioassay  methods  for 

acute  toxicity.    Water  Res.,  3  :  793-821. 

SPRAGUE,  J.  B.,  1971.    Measurement  of  pollutant  toxicity  to  fish.    III.  Sublethal  and  "safe"  con- 
centrations.  Water  Res.,  S  :  245-266. 
STUMM,  W.,  AND  J.  J.   MORGAN,   1970.     Aquatic  chemistry.    Wiley-Interscience,   New   York, 

New  York.  583  pp. 
TOWLE,  D.  W.,  G.  E.  PALMER,  AND  J.  L.  HARRIS  III,  1976.     Role  of  gill  Na+  +  Independent 

ATPase   in  acclimation  of  blue  crabs    (Callinectes  sapidns)    to   low   salinity.   /.   E.vp. 

Zool..  196:  315-322. 
TRUCHOT,  J.  P.,  1975.     Blood  acid-base  changes  during  experimental  emersion  and  reimmersion 

of  the  intertidal  crab,  Carcinus  maenas  (L.).    Rcspir.  Physio!.,  23:  351-360. 
TRUSSELL,   R.    P.,   1972.     The   percent   un-ionized   ammonia   in   aqueous   ammonia   solutions   at 

different  pH  levels  and  temperatures.     /.  Fish.  Res.  Board  Can..  29:   1505-1507. 
WALDBAUER,  G.  P.,   1968.     The  consumption  and  utilization  of  food  by  insecta.    Adv.  Insect. 

Physiol,  5 :  229-288. 

WARREN,  K.  S.,  1962.    Ammonia  toxicity  and  pH.  Nature.  195  :  47-49. 
WEILAND,  A.  L.,  AND  C.  P.  MANGUM,  1975.     The  influence  of  environmental  salinity  on  hemo- 

cyanin  function  in  the  blue  crab,  Callinectes  sapidus.    J.  E.rp.  Zool.,  193  :  265-273. 
WHITFIELD,  M.,   1974.     The  hydrolysis  of  ammonium  ions  in  sea  water — a  theoretical   study. 

/.  Mar.  Biol.  Assoc.  U.K.,  54 :  565-580. 

WICKINS,  J.  F.,   1976.     The  tolerance  of  warm-water  prawns  to   recirculated  water.    Aqua- 
culture,  9:  19-37. 
WUHRMANN,  V.  K.,  AND  H.  WORKER,  1948.    Beitrage  zur  toxikologie  der  fische.    II.  Experi- 

mentelle   untersuchungen   liber   die   ammoniak — und   blausaurevergiftung.     Schivciz.   Z. 

Hydro!.,  11:  210-244. 


KVtVrence:  Hiol.  />'»//..  154:  .\2-4f>.  (February. 


DESCRIPTION'S    OF    THE    LARVAE    OK    STICH  ASTER    AUSTRALIS 

(VERRILL)  kND  COSCINASTERIAS  CALAMARIA  (GRAY)  (  ECHINO- 

DERAI  ATA  :  ASTEROIDEA  )  FROM  XEW  ZEALAND, 

(  )I',TAINED  FROM  LABOKAT(  )k  V  CULTURE 

M.  F.  BARKER 

Department  of  Xm>lo<iy.  I  'nircrsity  of  Auckland,  Private  Bag,  Auckland,  AYzt1  Zealand 

In  asteroids,  as  in  other  classes  of  echinoderms.  there  are  two  main  types  of 
development.  Sonic  .species  have  indirect  development  with  a  free-swimming  larva 
which  may  feed  (planktotrophic  )  or  not  feed  (lecithotrophic)  in  the  plankton  ;  other 
species  have  direct  development,  with  no  free  stage,  and  eggs  which  may  lie  brooded 
liv  the  female  starfish. 

In  species  with  indirect  development  and  planktotrophic  larvae,  the  first  feeding 
stage  is  termed  a  hipinnaria.  This  generally  develops  into  a  brachiolaria  larva 
which,  at  the  end  of  development,  attaches  to  the  substratum  and  undergoes  meta- 
morphosis into  the  juvenile  starfish.  In  lecithotrophic  forms  and  those  with  direct 
development  and  even  in  some  species  with  planktonic  larvae,  this  sequence  may  be 
shortened  by  the  omission  of  larval  stages. 

By  the  early  part  of  this  century  a  number  of  investigators  had  attempted  to 
rear  different  asteroid  species  through  these  larval  stages  to  metamorphosis,  with 
varying  degrees  of  success.  Directly  developing  species  often  completed  develop- 
ment, but  planktotrophic  species  have  proved  more  difficult  to  rear.  This  early 
work  has  been  reviewed  by  Hyman  (1955  ). 

Since  these  early  studies,  there  have  been  few  detailed  investigations  of  asteroid 
development,  especially  of  species  with  indirect  development  and  planktotrophic 
larvae.  Chia  (1968)  has  described  the  development  of  the  brooding  starfish 
Lcptastcrias  lie.ractis  <  Stimpson)  ;  Birkeland,  Chia  and  Strathmann  (1971),  the 
development  of  nonfeeding  larvae  of  Mcdiastcr  ac<iiiulis  Stimpson  :  Komatsu  ( 1975), 
the  development  of  the  nonfeeding  larvae  of  Astropcctcn  lutcspinosns  Meissner; 
and  Atwood  (1973).  the  development  of  Echinastcr  echinophoruj  (Lamarck).  Of 
the  species  with  planktotrophic  larvae.  Greer  (1(<(>-?)  has  described  the  develop- 
ment of  Pycnopodla  liclimitlioidcs  (Brandt)  ;  Oguro,  Komatsu  and  Kano  (1976), 
the  development  of  Astropcctcn  sen  Darius  Valenciennes;  and  Henderson  (1969), 
Henderson  and  Lucas  (  1^71  )  and  Yamaguchi  (1973),  the  development  of  Acan- 
tlnistcr  planci  (L.).  Yamaguchi  (  1973)  has  also  included  some  information  on  the 
development  of  Linckia  hici'lijata  (L.)  and  Cidclta  noi'ncf/nincac  Muller  and 
Troschel;  and  Crump  (  l('(>'h  has  described  some  aspects  of  the  development  of  the 
Xew  Zealand  starfish  Patiriclla  rc</nlaris  (\Terrill).  Strathmann  (1971)  has 
reared  the  larvae  of  Patiria  ininntta  ( l->randt  )  and  I:rastcrius  troschcli  (  Stimpson) 
part  way  through  development  and  the  larvae  of  Lnidia  joliolata  Grube,  Pisastcr 
ochniccHs  (Brandt)  and  Pycnopodla  liclianthoidcs  through  to  metamorphosis,  but 
he  did  not  describe  the  larval  stages. 

Coscinastcrias  cahvnaria   ((iray")   is  widely  distributed  within  the  Indo-Pacific 

32 


LARVAL  DEVELOPMENT  OF  TWO   STARFISHES 

and  is  probably  the  most  common  forcipulate  starfish  in  Xe\v  Zealand.  Sfu'liastcr 
australis  (Yerrill)  is  endemic  to  Xe\v  Zealand  and  is  common  on  the  exposed  west 
coast  of  both  the  north  and  south  islands. 

Except  for  the  early  bipinnaria  of  C.  cakiinaria.  which  Morteiwn  (1921)  and 
Crump  (  1969)  obtained  from  in  I'itro  fertilized  eggs,  the  larval  stages  of  C.  cohi- 
nniria  and  S.  uustnilis  are  unknown.  This  paper  reports  on  the  methods  used  to 
rear  the  larvae  of  both  species  in  the  laboratory  and  the  larval  stages  are  described, 
Questions  of  breeding,  settlement  and  post-larval  development  will  be  considered  in 
later  publications. 

[MATERIALS  AND  METHODS 

Adult  starfish  of  Sticlnisfcr  australis  or  Coscinasterias  caluiuaria  were  collected 
from  Maori  Kay,  on  the  west  coast  of  Auckland,  during  the  breeding  season  from 
August  to  February.  The  ovaries  were  removed  in  the  laboratory  and  treated  with 
a  dilute  solution  (  1()"5  M)  of  1-methyladenine  (  Kanatani,  1969)  to  obtain  fertilizable 
eggs.  A  dilute  sperm  solution  was  prepared  from  small  pieces  of  mature,  excised 
testis.  Immobile  sperm  were  treated  with  a  solution  of  EOT  A  in  sea  water  to 
increase  their  activity.  All  sea  water  used  was  filtered  with  diatomaceous  earth  and 
glass  fiber  filters,  to  remove  all  extraneous  matter  over  1  /xm  diameter.  All  glass- 
ware used  was  washed  in  Pyroneg  cleaning  solution  and  soaked  in  dilute  sodium 
hypochlorite  solution  before  use. 

Eggs  obtained  by  the  above  method  were  washed  in  clean,  filtered  sea  water  and 
placed  in  a  500  ml  beaker;  a  few  drops  of  sperm  solution  were  added  and  allowed 
to  remain  with  the  eggs  for  5-10  minutes.  Eggs  were  then  filtered  off  in  bolting 
cloth  and  washed  in  several  changes  of  clean  sea  water  and  placed  in  5  liter  pyrex 
beakers  at  a  concentration  of  approximately  10  per  ml  of  water.  Beakers  were 
kept  under  natural  light  conditions  at  a  temperature  20°  C  :  1°  C  in  a  water  bath 
or  a  temperature-controlled  room. 

It  appears  that  some  degree  of  water  movement  is  important  for  the  successful 
culture  of  planktotrophic  asteroid  larvae.  Henderson  and  Lucas  (1971)  reared 
Acantliastcr  planci  larvae  in  culture  vessels  held  in  shaking  water  baths,  and  Gem- 
mill  (1914)  found  that  gentle  water  movements  were  necessary  when  rearing 
Asterias  rnbcns  L.  In  the  present  study,  cultures  were  kept  stirred  with  slowly 
revolving  paddles  at  10  rpm.  It  is  possible  that  such  movements  help  keep  food 
organisms  suspended  in  the  water  and  would  also  prevent  the  larvae  from  con- 
gregating on  the  bottom  of  the  culture  vessel  where  they  may  be  subject  to  greater 
risk  of  bacterial  infection. 

For  the  first  two  days,  developing  embryos  were  filtered  off  with  bolting  cloth 
every  12  hours  and  the  water  replaced.  In  addition,  an  antibiotic  solution  Crysta- 
mycin  (Glaxo  Lab. )  was  added  to  the  cultures.  This  was  added  at  a  concentration 
of  25  International  L'nits  benzyl  penicillin  sodium  and  0.25  X  10~4  g  streptomycin 
sulphate  per  ml  SW  of  the  cultures.  After  24  hours,  the  paddle  was  removed  and 
any  nondeveloping  eggs  allowed  to  sink  to  the  bottom  of  the  beaker.  By  this  stage 
most  developing  embryos  were  at  a  free  swimming  blastula  stage,  maintaining 
themselves  near  the  surface  of  the  beaker.  'They  were  filtered  off  and  placed  in  fre>h 


34  \|.  F.  BARKER 

culture  beakers  at  a  concentration  of  1  per  4  ml  of  sea  water.  Algal  food  was  added 
to  the  cultures  three  days  after  fertilization  for  C.  calamaria  and  four  days  after 
fertilization  for  S.  anstralis  larvae. 

Every  one  or  two  days,  samples  of  larvae  were  removed  with  a  pipette  to  check 
the  stage  of  development.  These  were  preserved  in  Bouins  or  neutralized  4% 
formalin. 

Every  two  days  the  water  was  changed  and  fresh  food  and  antibiotic  added.  On 
days  when  the  water  was  not  changed,  a  sample  of  water  was  taken  from  cultures 
and  the  concentration  of  algal  food  cells  present  was  determined  with  a  Coulter 
counter  (mean  of  three,  0.5  ml  samples).  Fresh  algae  were  added  to  replace  those 
cells  removed  by  feeding  larvae. 

Algal  foods  consisted  of  unialgal  (but  not  bacteria-free)  cultures  of  the  flagel- 
lates Dnnaliclla  pr'nnolccta  Butcher  and  IsocJirysis  galbana  Parke  and  the  diatom 
Phaeodactylum  tncornutum  Bohlin.  These  were  grown  in  modified  Guillard's 
medium  (Lanigan,  1972)  under  constant  illumination  from  overhead  strip  lights 
and  at  a  constant  temperature  of  20°  C.  Samples  of  algae  to  be  used  as  food  were 
•centrifuged  at  10,000  rpm  for  5  min ;  then  the  medium  was  poured  off,  and  the 
algae  were  resuspended  in  filtered  sea  water.  One  ml  samples  of  this  solution  were 
diluted  with  50  ml  filtered  sea  water,  and  the  cell  concentration  determined  with 
a  Coulter  counter  (mean  of  three,  0.5  ml  samples). 

The  concentration  and  type  of  algal  food  appears  to  be  important  in  the  culture 
of  starfish  larvae.  Early  in  this  study  cultures  were  fed  very  high  concentrations 
of  algal  species,  often  in  excess  of  50,000  cells/ml  in  the  culture  vessels.  This 
proved  detrimental,  particularly  in  the  case  of  S.  aiistralis,  and  few  larvae  in  these 
cultures  completed  development  and  cultures  had  a  high  percentage  of  deformed 
larvae  present.  The  stomachs  of  larvae  were  always  packed  with  food  but  much  of 
the  faeces  contained  intact  and  undigested  algal  cells  and  in  many  cases  the  intes- 
tine appeared  dark  and  swollen  with  a  mass  of  consolidated  faecal  matter.  It  was 
not  until  food  concentrations  were  reduced  to  the  levels  quoted  below  that  a  high 
percentage  of  normal  larvae  completed  development.  Although  IsocJirysis  galbana 
was  also  used  as  a  food,  including  mixtures  of  /.  (jalbana  with  D.  f>riinolccta  and 
P.  tncornutum,  the  most  satisfactory  food  for  C.  calamaria  larvae  was  Dunaliclla 
priinolccta  and  for  S.  anstralis,  Phaeodactylum  triconnitiim.  Although  other 
species  were  digested  and  some  growth  occurred,  in  very  few  cases  was  development 
completed  and  in  many  cultures  a  high  percentage  of  larvae  were  deformed. 

Algal  food  species  were  added  to  cultures  at  concentrations  of  8000  cells  per  ml 
of  culture  of  D.  priniolccta  and  10,000-12,000  cells  per  ml  for  /.  galbana  and  P. 
tricornutum.  In  cases  when  mixed  algae  were  fed,  the  combined  concentration  of 
cells  was  kept  at  approximately  10,000  cells  per  ml. 

RESULTS 
Spaivnlng  and  fertilization 

In  C.  calamaria,  ovaries  respond  to  treatment  with  1-methyladenine  from  early 
to  late  in  the  breeding  season  (July-February).  In  ,S\  anstralis  ovaries  only  re- 
spond to  this  treatment  late  in  the  breeding  season  when  the  gonad  is  very  ripe. 


LARVAL  DEVELOPMENT  OF  TWO   STARFISHES  35 

Kanatani  (1969)  also  found  1-methyladenine  to  be  more  effective  in  stimulating 
spawning  of  Astropecten  anrantiacns  (Tiedemann)  later  in  the  breeding  season. 

From  60  to  120  minutes  (for  5".  anstralis}  and  40  to  60  minutes  (for  C. 
calamaria)  after  being  placed  in  this  solution  the  ovary  wall  contracts,  and  mature 
ova  are  released  from  the  oviduct  and  from  any  breaks  in  the  ovary  wall.  Mature 
ova  are  spherical,  almost  translucent  and  the  color  varies  from  cream  to  light  yellow 
in  S.  anstralis  and  from  red  to  orange  in  C.  calauiaria.  They  are  enclosed  in  a  jelly 
layer  10-15  ju,m  thick,  and  in  5".  anstralis  have  a  total  diameter  of  120-140  /xm  and 
in  C.  calauiaria,  a  diameter  of  140-160  /xm.  On  spawning,  the  germinal  vesicle  has 
either  already  disappeared  or  is  breaking  down  rapidly. 

The  first  indication  of  fertilization  is  a  lightening  in  color  of  the  surface  of  the 
egg,  especially  around  the  plasma  membrane  in  the  region  where  the  perivitelline 
space  will  form  (Fig.  la).  Within  30  seconds  to  1  minute  the  fertilization  mem- 
brane begins  to  lift,  probably  at  the  point  of  sperm  entry.  Elevation  continues 
rapidly  around  the  egg,  normally  being  complete  after  2-5  minutes,  and  a  narrow 
perivitelline  space  of  3-5  ^m  is  apparent.  The  fertilization  membrane  continues  to 
expand  until,  after  5-6  minutes,  there  is  a  gap  between  the  plasma  and  fertilization 
membranes  of  15-20  /mi.  Within  5-10  minutes  the  first  polar  body  is  extruded 
(Fig.  2a,  Ib)  and  the  second  polar  body  follows  1  to  1.5  hours  later.  At  this  stage 
the  fertilized  egg  lies  free  within  the  fertilization  membrane,  the  perivitelline  space 
is  about  15-20  //m  in  diameter,  and  the  jelly  coat  has  dissolved. 

Ewibry  agenesis 

As  is  typical  in  echinoderms,  cleavage  is  radial  and  holoblastic ;  however,  the 
rate  of  embryonic  development  varies  greatly  for  eggs  from  the  same  animal 
fertilized  at  the  same  time.  The  times  given  in  the  following  description  of  the 
development  of  S.  anstralis  and  C.  calauiaria  are  the  average  time  after  fertilization, 
for  a  particular  stage  to  be  reached,  for  the  majority  of  embryos  at  20°  C. 

From  1-2  hours  (after  fertilization),  the  first  cleavage  occurred  (Fig.  2b),  and 
after  3  hours  the  second  and  third  cleavages  were  complete.  At  5-8  hours  develop- 
ing embryos  were  at  the  64-128  blastomere  stage,  the  blastocoel  now  becoming 
obvious.  At  16-18  hours  the  ciliated  coeloblastula  was  revolving  within  the 
fertilization  membrane  (Fig.  2e),  and  at  20-22  hours  the  fertilization  membrane 
ruptured  releasing  the  developing  embryo  (Fig.  le).  At  this  stage  typical  embolic 
imagination  was  commencing  at  the  vegetal  pole.  At  30-36  hours  gastrulation  was 
Avell  advanced,  and  mesenchyme  formation  at  the  tip  of  the  advancing  archenteron 
was  occurring  (Figs.  2f  and  If).  At  45-46  hours  embryos  were  at  the  late  gastrula 
stage,  imagination  being  nearly  complete.  Left  and  right  enterocoels  were  cut 
off  from  the  tip  of  the  archenteron,  and  stomodaeal  invagination  was  well  advanced. 
At  55-60  hours  stomodaeum  formation  was  complete,  the  gut  was  present  as  a 
narrow  tube,  but  the  bulbous  stomach  was  not  yet  differentiated.  The  larva  was 
now  becoming  dorsoventrally  flattened  and  was  beginning  to  assume  the  shape  of 
the  bipinnaria,  0.4  mm  in  total  length  (Fig.  2g).  The  left  and  right  enterocoels 
were  lying  beside  the  gut,  the  left  having  put  out  a  narrow  evagination  to  the  dorsal 
surface  to  form  the  hydropore.  At  70-80  hours  embryonic  development  was  com- 
plete, with  the  stomach  fully  formed. 


M.  F.  BAKKKK 


LARVAL  DEYKLOPMKXT  OF  TWO   STARFISHES  37 

Dcrclo  fluent  of  the  jccdhi;/  lari'a 

Although  there  are  minor  morphological  differences,  larval  development  in  5". 
australis  and  C.  cahunaria  is  very  similar  and  the  following  description  applies  to 
hoth  species.  As  with  emhryonic  development,  the  rate  of  development  of  feeding 
larvae  in  a  particular  culture  shows  considerable  variation,  and  the  times  given  for 
a  particular  stage  to  he  reached  are  the  average  time  after  fertilization  for  the 
majority  of  larvae  at  20°  C.  C.  cahunaria  has  a  slightly  faster  rate  of  development 
than  .V.  australis,  and  this  difference  becomes  more  pronounced  as  development 
proceeds.  Sizes  given  are  the  total  length  of  the  larva. 

The  early  bipinnaria  stage  (Fig.  Ig,  2h  )  was  reached  4  days  after  fertilization 
and  the  larvae  of  both  C.  cahunaria  and  .V.  ausfralis  were  0.4-0.5  mm  in  length. 
Their  major  ciliated  tracts,  consisting  of  a  preoral  loop  and  postoral  and  lateral 
bands,  were  complete  and  the  larvae  were  feeding.  At  8  days  (C.  cahunaria,  0.8— 
0.9  mm;  6".  australis.  0.9-1.0  mm)  the  left  and  right  enterocoels  were  extending 
anteriorly  and  posteriorly.  The  ciliated  tracts  were  becoming  more  complex  and 
expanding  in  the  regions  where  the  processes  would  form.  At  10-11  days  (C. 
cahunaria,  1.2-1.3  mm.  Fig.  Ih,  3a ;  S.  australis,  1.0-1.1  mm.  Fig.  2i.  4a )  the 
enterocoels  were  continuing  to  grow  anteriorly  and  posteriorly,  the  left  often  slightly 
better  developed  than  the  right.  Posteriorly  the  ventral  horn  of  the  left  posterior 
coelom  was  beginning  to  form.  At  14  days  for  C.  cahunaria  (  1.7  mm  )  and  K>  days 
for  S.  australis  (1.2—1.3  mm)  the  preoral  (paired),  postoral  (paired),  anterior 
dorsal  (paired),  median  dorsal  (single),  posterior  lateral  (paired)  and  posterior 
dorsal  (paired)  processes  were  beginning  to  form.  Left  and  right  enterocoels  were 
joined,  anterior  to  the  stomodaeum  and  were  growing  into  the  median  dorsal  pro- 
cess (mdp).  Posteriorly  the  enterocoels  [termed  the  left  posterior  coelom 
(Ipc)  and  the  right  posterior  coelom  ( rpc ) .  (Gemmill,  1914)  |  were  beginning 
to  surround  the  stomach  and  intestine.  The  ventral  horn  of  the  Ipc  was  growing 
toward  the  rpc,  ventral  to  the  stomach  and  anterior  to  the  intestine;  and  the  Ipc 
was  growing  around  the  posterior  region  of  the  stomach  and  intestine.  In  C. 
calaiuaria  a  dorso-ventral  cleft  was  forming  at  the  end  of  the  mdp. 

The  late  bipinnaria  stage  was  reached  at  16  davs  for  C.  calaiuaria  (1.8-1.9  mm) 
and  at  20  days  for  S.  australis  (  1.6-1.7  mm.  Fig.  2j  ).  This  stage  was  marked  by 
the  lengthening  processes  which  in  S.  australis  were  becoming  pigmented  (brown) 
at  the  tips.  The  anterior  coelom  was  advancing  well  up  the  mdp  and  in  the  most 
advanced  specimens  was  growing  out  as  two  buds  in  the  region  where  the  posterior 
brachiolar  arms  would  form.  The  posterior  coelom  had  almost  completed  develop- 
ment around  the  stomach  and  intestine.  A  septum  was  forming  separating  the 


FIGURE  1.  Development  of  Coscinasterias  cttldiimriit  (times  given  in  a-k  are  the  period 
after  fertilization)  :  a)  egg  immediately  after  fertilization,  note  the  incipient  fertilization  mem- 
brane around  the  periphery  of  the  plasma  membrane  (arrowed)  :  b)  5  minutes,  note  the  presence 
of  a  germinal  vesicle  (arrowed)  ;  c)  2  hours,  4  blastomere  stage;  d)  4  hours  early  hlastula ;  e) 
20-21  hours,  the  fertilization  membrane  ruptures  releasing  the  ciliated  coeloblastula  ;  f )  3f>  hours, 
advanced  gastrula ;  g)  4  days,  ventral  view  of  early  bipinnaria;  hi  11  days,  bipinnaria,  dor>al 
view;  i)  18  days,  ventral  view  of  early  hrachiolaria  ;  j  )  20  days,  brachiolaria,  side  vk\\  ;  k)  24 
days,  side  view  of  well  developed  brachiolaria;  and  1  )  juvenile  starfish  <>  days  after  attachment 
to  the  substratum.  Scale  is  0.05  mm  in  a-f ;  0.1  mm  in  ,i>-j  ;  and  0.5  mm  in  k-1. 


M.  F.  BARKER 


LARVAL  DEVELOPMENT  OF  TWO  STARFISHES  39 

left  posterior  from  the  left  middle  coelomic  region,  and  a  partial  septum  separating 
the  right  posterior  from  right  middle  coelomic  region  was  also  forming.  In  S. 
australis  a  dorso-ventral  cleft  was  forming  at  the  tip  of  the  mdp. 

The  early  brachiolaria  stage  [18  days  for  C.  calamana  (2.1-2.2  mm,  Fig.  li) 
and  24  days  for  6".  australis  (2.1-2.3  mm)]  was  marked  by  the  formation  of  rudi- 
mentary brachiolar  arms,  as  outgrowths  at  the  bases  of  the  preoral  processes  and 
on  the  ventral  tip  of  the  mdp.  The  anterior  coelom  was  bulging  into  these,  and 
the  adhesive  disc  was  just  becoming  apparent  on  the  ventral  sides  of  the  mdp, 
between  the  three  brachiolar  arms.  The  processes  were  now  quite  elongate  (in  5\ 
oits trails  with  light  brown  pigmentation  at  their  terminal  ends).  At  20  days  for 
C.  calamana  (2.5  mm,  Fig.  Ij)  and  27  days  for  5".  australis  (2.4-2.5  mm)  the 
brachiolar  arms  were  lengthening  and  adhesive  papillae  were  forming  on  their  tips. 
The  adhesive  disc  was  now  well-developed  and  five  hydrocoel  lobes  were  develop- 
ing on  the  left  posterior  coelom.  Posteriorly,  the  starfish  primordium  was  ap- 
parent as  five  rounded  outgrowths.  At  24  days  for  C.  calamana  (3.2-3.3  mm, 
Fig.  Ik,  3b)  and  31  days  for  S.  australis  (3.2  mm,  Fig.  2k,  4b)  the  processes  were 
very  elongate  and  the  brachiolar  arms  had  also  elongated  and  were  pigmented 
yellow-brown  in  both  5".  australis  and  C.  calamana.  Each  terminated  in  a  crown  of 
adhesive  papillae.  The  five  hydrocoel  lobes  were  now  more  symmetrically  arranged 
on  the  oral  side  of  the  starfish  primordium.  Rudimentary  spines  and  ossicles  were 
forming  on  what  would  become  the  aboral  surface  of  the  yellow7  brown  primordium. 
The  primordium  is  dorso-ventrally  orientated  so  that  the  oral  side  of  the  developing 
juvenile  starfish  is  on  the  righthand  side  in  Figures  3b  and  4b. 

At  the  late  brachiolaria  stage  [27  days  for  C.  calamana  (3.6-3.7  mm)  and  38 
days  for  S.  australis  (3.6-3.7  mm)],  the  starfish  primordium  was  well-developed 
with  pronounced  ossicles  and  spines.  In  culture  vessels  larvae  were  often  swimming 
near  the  bottom  and  if  suitable  substrata  were  presented  would  attach  and  undergo 
metamorphosis. 

Metamorphosis 

The  process  of  metamorphosis  is  very  similar  in  C.  calamana  and  6\  australis 
and  the  following  description  applies  to  both  species.  As  with  development  of  the 
feeding  larvae,  metamorphosis  proceeded  at  a  slightly  faster  rate  in  C.  calamana 
than  in  S.  australis.  The  following  times  are  for  the  different  stages  of  develop- 
ment after  permanent  attachment  to  the  substratum  by  the  adhesive  disc. 

From  0-6  hours  the  anterior  portion  of  the  larva  bearing  the  ciliated  processes, 
brachiolar  arms,  larval  mouth,  etc.,  gradually  contracted  and  at  the  same  time  under- 

FIGURE  2.  Development  of  Stichastcr  australis  (times  given  in  a-k  are  the  period  after 
fertilization)  :  a)  egg  at  6  minutes,  note  the  presence  of  a  germinal  vesicle  (arrowed)  ;  b) 
1.5  hours,  the  first  cleavage  is  in  progress;  c)  3  hours,  8  blastomere  stage;  d)  6  hours,  early 
blastula ;  e)  16  hours,  the  coeloblastula  is  revolving  within  the  fertilization  membrane;  f)  36 
hours,  advanced  gastrula;  g)  60  hours,  ventral  view  of  very  early  bipinnaria — the  gut  is 
present  as  a  narrow  tube,  but  the  bulbous  stomach  is  not  yet  differentiated;  h)  4  days,  ventral 
view  of  early  bipinnaria;  i)  11  days,  ventral  view  of  bipinnaria;  j)  20  days,  ventral  view  of 
late  bipinnaria;  k)  31  days,  well  developed  brachiolaria;  and  1)  juvenile  starfish  6  days  after 
attachment  to  the  substratum.  Scale  is  0.05  mm  in  a-g ;  0.1  mm  in  h-j  ;  and  0.5  mm  in  k-1. 


40 


M.  F.  BARKER 


re 


vh 


FIGURE  3.  Larval  stages  of  Coscinasterias  calamaria:  a)  bipinnaria,  11  days  after  fertiliza- 
tion; b)  well-developed  brachiolaria  24  days  after  fertilization.  Abbreviations  are:  ad,  adbesive 
disc;  adp,  anterior  dorsal  process;  ap,  adbesive  papillae;  ba,  brachiolar  arms;  hcl,  bydrocoel 
lobes ;  by,  hydropore ;  bye,  hydropore  canal ;  le,  left  enterocoel ;  Ipc,  left  posterior  coelom ;  mdp, 
median  dorsal  process ;  pdp,  posterior  dorsal  process ;  pip,  posterior  lateral  process ;  pr, 
primordium ;  prp,  preoral  process ;  ptp,  postoral  process ;  re,  right  enterocoel ;  and  vh,  ventral 
horn  of  left  posterior  coelom.  Scale  is  0.5  mm. 


went  torsion,  bringing  the  oral  side  of  the  developing  starfish  primordium  closer  and 
parallel  to  the  substratum.     At  12  hours  the  ciliated  processes  were  tightly  coiled 


LARVAL  DEVELOPMENT  OF  TWO   STARFISHES 


41 


and  partially  resorbed.  The  "larval  notch"  (gap  between  ray  1  and  5  where  the 
larval  body  (now  attachment  stalk)  was  formed)  was  becoming  smaller  and  yellow 
pigmentation  of  the  primordium  was  darkening.  At  23  hours  shortening  of  the 
attachment  stalk  was  continuing.  The  hydmcoel  lobes  had  expanded  somewhat  and 
were  lying  more  centrally  on  the  oral  side  of  the  primordium.  In  C.  calaniaria 
division  of  each  of  the  hydrocoel  lobes  into  two  pairs  of  tube  feet  had  commenced 


rpc 


\.nlr j|   \ ie« 


mdp 


Wnlrjl  \  ir» 


FIGURE  4.  Larval  stages  of  Stichaster  australis:  a)  bipinnaria,  10  days  after  fertilization; 
b)  well-developed  brachiolaria,  31  days  after  fertilization.  Abbreviations  are:  ad,  adhesive 
disc ;  adp,  anterior  dorsal  process ;  ap,  adhesive  papillae ;  ba,  brachiolar  arms :  hcl,  hydrocoel 
lobes ;  hy,  hydropore ;  hyc,  hydropore  canal ;  le,  left  enterocoel ;  Ipc,  left  posterior  coelom ;  mdp, 
median  dorsal  process;  pdp,  posterior  dorsal  process;  pip,  posterior  lateral  process;  pr,  primor- 
dium; prp,  preoral  process;  ptp,  postoral  process;  re,  right  enterocoel;  rpc,  right  posterior 
coelom;  and  vh,  ventral  horn  of  left  posterior  coelom.  Scale  is  0.5  mm. 


42  M.  F.  BARKER 

but  discrete  podia  were  not  yet  obvious.  At  27  hours  the  ciliated  processes  were 
fully  resorbecl.  The  5  rays  of  the  developing  disc  were  symetrically  arranged,  and 
the  larval  notch  was  no  longer  obvious.  Aborally  the  skeletal  plates  were  starting 
to  assume  their  adult  arrangement  of  a  central  and  5  primary  interradial  and  5 
terminal  plates.  Each  bore  1  or  2  small  spines.  At  36  hours  the  juvenile  starfish 
was  pulled  down  close  to  the  substratum,  the  ossicles  were  becoming  larger  and 
were  covering  a  greater  area  of  the  aboral  surface.  In  the  development  of  the 
water  vascular  system,  in  S.  anstralis  each  hydrocoel  lobe  was  dividing  into  two 
pairs  of  podia.  In  C.  calamaria  development  had  proceeded  further  and  small 
distinct  podia  and  developing  radial  canals  could  be  seen.  At  52  hours  separate 
podia  and  radial  canals  were  obvious  in  S.  anstralis.  In  C.  calamaria  podia  were  well 
formed  and  were  attached  to  the  substratum,  and  terminal  tentacles  had  also  devel- 
oped. At  80  hours  in  S.  anstralis  podia  were  now  attached  to  the  substratum  and 
terminal  tentacles  were  present.  In  C.  calamaria  a  red  eyespot  had  formed  at  the 
base  of  each  terminal  tentacle.  At  4  days  red  eyespots  were  forming  at  the  base 
of  each  terminal  tentacle  in  S.  anstralis.  In  C.  calamaria  the  obvious  external 
changes  of  metamorphosis  appeared  complete.  Podia  were  being  used  to  move  the 
juvenile  against  the  restraining  attachment  stalk  which  was  becoming  thinner.  At 
5  days  the  obvious  external  changes  of  metamorphosis  appeared  to  be  complete  in 
S.  anstralis,  and  the  podia  were  moving  the  juvenile  starfish  against  the  restraining 
attachment  stalk.  At  6  days  for  C.  calamaria  (Fig.  11)  and  at  6-7  days  for  S. 
anstralis  (Fig.  21)  the  attachment  stalk  ruptured  close  to  the  point  of  attachment  to 
the  substratum,  and  the  small  juvenile  starfish  assumed  free  life.  At  this  stage 
there  were  5  primary  rays,  with  4  podia  and  1  terminal  tentacle  per  ray.  The  total 
diameter  of  the  disc  was  0.9  mm.  The  adult  mouth  had  not  yet  formed.  At  10-12 
days  formation  of  the  adult  mouth  was  complete,  and  the  small  juvenile  starfish 
(0.95  mm  in  diameter)  commenced  feeding. 

DISCUSSION 

The  development  of  5".  anstralis  and  C.  calamaria  follows  closely  the  published 
descriptions  of  the  development  of  other  planktotrophic  asteroid  larvae.  The  most 
comprehensive  of  these  descriptions  is  that  given  by  Gemmill  (1914)  for  Asterias 
rnbcns.  C.  calamaria,  S.  anstralis  and  A.  rnbcns  are  all  members  of  the  family 
Asteriidae.  The  larvae  of  C.  calamaria  and  A.  rubens  are  very  similar;  S.  anstralis, 
on  the  other  hand,  while  showing  the  same  general  pattern  of  development  as  A. 
rubens  and  C.  calamaria,  exhibits  slight  differences. 

S.  anstralis  larvae,  for  example,  have  more  strongly  pigmented  and  shorter 
processes  and  shorter  and  more  rounded  brachiolar  arms  than  C.  calamaria  larvae. 
The  slower  growth  rate  may  reflect  differences  in  the  laboratory  culture  conditions, 
rather  than  inherent  differences  in  the  biology  of  the  two  species. 

The  main  structural  differences  which  larval  stages  of  C.  calamaria  and  S. 
anstralis  show  when  compared  to  Asterias  rnbcns  are :  the  former  species  do  not 
develop  a  dorsal  sac,  the  coelomic  epithelium  has  few  cilia,  and  there  appears  to  be 
no  movement  of  fluid  within  the  coelomic  cavities.  In  those  species  where  it  is 
present,  the  dorsal  sac  appears  to  exhibit  rhythmic  contractions,  and,  although  it 
lacks  an  inlet  or  an  outlet,  it  apparently  maintains  coelomic  circulation  within  the 


LARVAL  DEVELOPMENT  OF  TWO   STARFISHES  43 

larvae  by  passing  fluid  through  its  walls  (Gemmill,  1914).  The  observed  lack  of 
movement  of  the  coelomic  fluid  in  C.  calamaria  and  S.  australis  may  be  due  to  the 
absence  of  this  structure,  and  coelomic  circulation  in  these  species  would  appear  to 
be  unnecessary. 

Larvae  exhibiting  structural  abnormalities  have  been  described  by  several  authors 
(Gemmill,  1914;  MacBride,  1896;  and  Xewth,  1925).  A  double  hydropore,  single 
enterocoel,  differing  growth  rates  of  enterocoels,  and  absence  of  the  median  dorsal 
process  are  common  variations  noted  in  the  literature. 

In  the  present  study  irregularities  in  development  were  found  for  both  C. 
calamaria  and  S.  australis.  Variations  in  the  growth  rates  of  enterocoels  (i.e.,  one 
enterocoel  expanding  much  faster  than  the  other)  were  common  in  cultures  of  both 
species  and  would  seem  to  be  normal  growth  variations.  In  C.  calamaria  other 
irregularities  in  larval  development  were  only  rarely  encountered.  In  some  S. 
australis  cultures,  however,  the  absence  of  the  median  dorsal  process  or  of  one  or 
other  of  the  lateral  processes  occurred  in  up  to  80%  of  the  larvae.  In  other  cultures, 
almost  100%  of  the  larvae  would  develop  normally.  Abnormal  larvae  seemed  to 
develop  most  commonly  in  those  cultures  fed  on  species  other  than  D.  primolecta 
for  C.  calamaria  and  P.  tricornutum  for  6".  australis,  or  in  cultures  fed  a  particularly 
high  concentration  of  food.  It  would  seem,  therefore,  that  abnormal  larvae  are  a 
result  of  unsuitable  culture  conditions,  although  it  is  possible  some  abnormalities 
have  a  genetic  origin. 

Culture  conditions  may  also  contribute  to  the  wide  variation  in  growth  rates  of 
larvae  within  a  particular  culture  noted  above.  Similar  variations  in  growth  rates 
also  occur  in  Pycnopodia  JicliantJwidcs  (Greer,  1962)  and  Asterias  nibens  (Gem- 
mill,  1914)  and  are  probably  quite  normal  in  planktotrophic  larvae.  In  contrast, 
larvae  of  Mediaster  eqnalis,  a  species  with  direct  development,  show  little  varia- 
tion in  size  (Birkeland  ct  a/.,  1971),  and  it  seems  likely  that  lecithotrophic  species, 
with  their  yolk  reserves,  have  a  much  more  synchronous  development. 

The  internal  reorganization  of  tissues  at  metamorphosis  is  complex  and  has 
been  described  fully  by  Gemmill  (1914).  The  external  changes  that  occur  at 
metamorphosis  in  5\  australis  and  C.  calamaria  parallel  those  described  by  Gemmill 
for  Asterias  rubcns,  except  that  in  A.  rubens  three  pairs  of  podia  are  formed  per 
ray  at  the  completion  of  metamorphosis,  while  in  S.  australis  and  C.  calamaria  two 
pairs  are  formed.  This  is,  however,  a  minor  point  and  is  unlikely  to  reflect  any 
major  differences  in  internal  structure.  It,  therefore,  seems  likely  that  meta- 
morphosis in  .S\  australis  and  C.  calamaria  follows  the  basic  pattern  described  by 
Gemmill  (1914)  and  further  detailed  description  is  unnecessary. 

Many  marine  invertebrates  have  a  planktonic  dispersal  phase  in  their  life  history, 
at  the  end  of  which  occurs  settlement  and  metamorphosis  into  the  adult.  Despite 
the  complex  structural  reorganization  that  occurs,  once  settlement  and  the  more 
obvious  changes  at  metamorphosis  commence,  they  are  usually  completed  in  a  com- 
paratively short  time.  For  example,  in  barnacle  cyprids  the  change  from  the  larval 
to  the  adult  form  may  be  complete  in  24  to  as  little  as  8  hours.  The  reasons  for 
this  are  fairly  obvious.  Demands  on  food  reserves  at  this  time  are  high  and  as 
changes  from  larval  life  to  adult  life  are  generally  accompanied  by  a  drastically 
altered  diet,  feeding  must  be  quickly  initiated  again.  Also,  at  this  time  the  larva  or 


44  M.  F.  BARKER 

early  juvenile  must  be  very  susceptible  to  preclation.  In  view  of  this,  it  would  ap- 
pear somewhat  remarkable  that  juvenile  starfish  do  not  break  free  from  the  sub- 
stratum until  six  days  after  attachment  by  the  brachiolaria.  In  starfish,  however, 
the  problem  of  food  reserves  is  largely  solved  by  resorption  and  re-utilization  of 
larval  structures  as  metamorphosis  proceeds,  the  juvenile  starfish  generally  being 
much  smaller  than  the  advanced  brachiolaria.  Although  it  would  appear  that  the 
delicate,  recently  attached  brachiolaria  or  partly  metamorphosed  juvenile  is  very 
vulnerable  to  predation,  there  is  some  evidence  that  the  production  of  toxic  sub- 
stances, possibly  saponins,  deters  predators.  Personal  observations  have  shown 
that  some  potential  predators,  such  as  polychaetes,  avoid  or  quickly  release  search- 
ing larvae  if  they  come  into  direct  contact,  and  Yamaguchi  (1975)  has  made 
similar  observations. 

Hyman  (1955)  noted  the  tendency  for  related  asteroid  species  to  hybridize, 
producing  puzzling  specimens.  Gemmill  (1914)  found  that  a  high  percentage  of 
oocytes  of  Astcrias  rubcns  were  fertilized  by  Marthasterias  cjlacialis  (L.)  sperm  and 
vice  versa,  and  a  number  of  these  oocytes  proceeded  to  the  blastula  or  gastrula  stage. 
He  also  found  that  fertilization  of  small  numbers  of  A.  rubcns  oocytes  could  be 
achieved  with  sperm  from  other  genera,  although  development  did  not  proceed 
past  the  early  cleavage  stages.  Gemmill  (1916)  also  cross  fertilized  Stichastrella 
rosea  (Miiller)  oocytes  with  Marthasterias  glacialis  sperm  and  found  that  develop- 
ment proceeded  normally  to  the  early  bipinnaria  stage  in  a  small  proportion  of  eggs. 
Lucas  and  Jones  (1976)  cross  fertilized  oocytes  of  Acanthaster  planci  and  A. 
brevispinus  Fisher  with  sperm  of  the  other  species  and  managed  to  rear  the  resulting 
larvae  to  adult  starfish. 

On  one  occasion,  when  ripe  individuals  of  both  S.  australis  and  C.  calainaria 
were  present  in  the  laboratory  attempts  were  made  to  fertilize  ova  of  the  one  species 
with  sperm  of  the  other.  With  S.  australis  oocytes  and  C.  calainaria  sperm  a  small 
number  of  oocytes  were  fertilized.  Irregular  cleavage  produced  a  deformed  blastula 
with  a  poorly  formed  blastocoel  after  22  hours,  and  development  did  not  proceed 
further.  It  was  found  that  a  higher  percentage  of  C.  calainaria  oocytes  were  fertil- 
ized with  S.  aiistralis  sperm,  and  although  the  subsequent  pattern  of  development 
was  somewrhat  irregular  in  most  oocytes,  in  a  few  development  proceeded  normally, 
and  a  well  formed  gastrula  resulted  36  hours  after  fertilization.  Unfortunately,  a 
shortage  of  culture  facilities  did  not  allow  continuation  of  this  experiment.  How- 
ever, the  development  of  some  apparently  normal  gastrulae,  plus  the  similar  mor- 
phology of  the  larvae  of  C.  calamaria,  A.  rubens  and  S.  australis,  all  members  of 
the  Asteriidae,  lends  support  to  the  suggestion  of  Oguro  ct  al.  (1976)  that  in  some 
asteroids,  developmental  features  are  related  to  the  systematic  position  of  the  species. 


I  wish  to  thank  the  University  Grants  Committee  for  the  support  of  a  post- 
graduate scholarship  and  Dr.  B.  A.  Foster  for  his  advice  and  for  critically  reading 
the  manuscript. 


LARVAL  DEVELOPMENT  OF  T\YO   STARFISHES  45 

SUMMARY 

1.  Methods   for  the   laboratory   rearing  of   larvae   of   the   starfishes   Stichaster 
australis  and  Coscinastcrias  calainaria  are  described. 

2.  Larval  development  in  S.  australis  and  C.  calainaria  is  very  similar,  although 
C.  calainaria  has  a  slightly  faster  rate  of  development.     Fertilized  eggs  develop 
through  a  bipinnaria  to  a  brachiolaria  stage.     Late  brachiolaria  larvae  were  present 
38  days   after   fertilization   in   S.   australis   and   27   days   after   fertilization   in    C. 
calainaria. 

3.  As  in  the  development  of  the  feeding  larvae,  the  process  of  metamorphosis 
is  very  similar  in  S.  australis  and  C.  calainaria.     The  time  from  attachment  to  the 
substratum  by  the  late  brachiolaria  larvae  to  the  completion  of  metamorphosis  of 
the  juvenile  starfish  is  6-7  days  in  S.  australis  and  6  days  in  C.  calainaria. 

4.  Unfavorable  culture  conditions  may  have  been  the  cause  of  abnormal  larvae 
found  in  some  cultures. 

5.  Larval  development  of  .S".  australis  and  C.  calainaria  resembles  closely  that 
of  other   starfish   species   with   indirect   development,   especially   Asterias  rubens. 
This  may  reflect  the  close  taxonomic  affinities  of  these  three  species. 

LITERATURE  CITED 

ATWOOD,  D.  G.,  1973.     Larval  development  in  the  asteroid  Echinastcr  cchinophorus.    BioL  Bull., 

144:  1-11. 
BIRKELAND,  C.,  F.  S.  CiiiA,  AND  R.  S.  STRATHMANN,  1971.     Development,  substrate  selection, 

delay  of  metamorphosis  and  growth  in  the  seastar  Mcdiaster  acqualis.  BioL  Bull.,  141  : 

99-108. 
CHIA,  F.  S.,  1968.  The  embryology  of  a  brooding  starfish,  Lcptastcrias  hcxactis   (Stimpson). 

Act  a  ZooL,  49:    321-364. 
CRUMP,  R.  G.,  1969.    Aspects  of  the  biology  of  some  New  Zealand  echinoderms.    Ph.D.  thesis. 

University  of  Otago,  Dunedin,  New  Zealand. 

GEMMILL,  J.  F.,  1914.     The  development  and  certain  points  in  the  adult  structure  of  the  star- 
fish Asterias  rubens  L.  Phil.  Trans.  Roy.  Soc.  London  Ser.  B,  205:  213-294. 
GEMMILL,  J.  F.,   1916.     Notes  on  the  development  of  the  starfishes  Asterias  glacialis  O.F.M., 

CribrcIIa  oculata   (Linck)    Forbes,  Stichaster  rosens   (O.F.M.)    Sars.  Proc.  ZooL  Soc. 

London.  39:  553-565. 
GREEK,  D.  L.,  1962.    Studies  on  the  embryology  of  Pycnopodia  helianthoides  (Brandt)   Stimpson. 

Pacific  Set.,  16:  280-285. 
HENDERSON,  J.  A.,   1969.     Preliminary  observations  on  the  rearing  and  development  of  Acan- 

thaster  planci   (L.)    (Asteroidea)    larvae.    Fish.  Notes  Queensland,  3:  69-75. 
HENDERSON,  J.  A.  AND  J.   S.  LUCAS,   1971.     Larval  development  and  metamorphosis  of  Acan- 

thaster  planci  (Asteroidea).    Nature,  232:  655-657. 
HYMAN,    L.    H.,    1955.      The   Invertebrates,    IV.   Echinodcrmata.     McGraw-Hill.    New   York, 

763  pp. 

KANATANI,  H.,  1969.     Induction  of  spawning  and  oocyte  maturation  by  1-methyladenine  in  star- 
fishes.   Exp.  Cell  Res.,  57 :  333-337. 
KOMATSU,   M.,   1975.     On  the  development  of  the  sea-star,  Astropccten   latespinosus  Meissner. 

BioL  Bull..  148:  49-59. 
LANIGAN,   K.   A.,    1972.   Nutrients   influencing   phytoplankton   growth   in   the   Jellicoe   Channel. 

.1/..SV.  thesis.  University  of  Auckland,  Auckland,  New  Zealand.  114  pp. 
LUCAS,  J.  S.,  AND  M.  M.  JONES,  1976.     Hybrid  crown-of-thorns  starfish  (Acanthastcr  planci  X 

A.  brcvispimts)   reared  to  maturity  in  the  laboratory.     Nature,  263:  409-412. 
MACBRIDE,  E.  W.,  1896.    The  development  of  Asterina  gibbosa.   Q.  J.  Micros.  Sci.,  38:  339-441. 
MORTENSEN,  J.,   1921.     Studies  of  the  development  and  larval  forms  of  echinoderms.    G.E.C. 

Gad,  Copenhagen,  216  pp. 


46  M.  F.  BARKER 

NEWTH,  H.  G.,  1925.  The  early  development  of  Astropecten  irrcgnlaris  with  remarks  on  du- 
plicity in  echinoderm  larvae.  /.  Microsc.  Sci.,  69:  519-554. 

OGURO,  C,  M.  KOMATSU  AND  Y.  T.  KANO,  1976.  Development  and  metamorphosis  of  the  sea- 
star  Astropecten  scoparius  Valenciennes.  Biol.  Bull.,  151  :  560-573. 

STRATHMANN,  R.  R.,  1971.  The  feeding  behaviour  of  planktotrophic  echinoderm  larvae:  mecha- 
nisms, regulation,  and  rates  of  suspension-feeding.  /.  E.vp.  Mar.  Biol.  Ecol.,  6:  109-160. 

YAMAGUCHI,  M.,  1973.  Early  life  histories  of  coral  reef  asteroids,  with  special  reference  to 
Acanthastcr  planci  (L.).  Pages  369-387  in  O.  A.  Jones  and  R.  Endean,  Eds.,  Biology 
and  geology  of  coral  reefs,  Vol.  2,  Biology  1.  Academic  Press,  New  York. 

YAMAGUCHI,  M.,  1975.     Coral-reef  asteroids  of  Guam.    Biotropica,  7 :  12-23. 


1 


Reference:  Blol.  Bull,  154:  47-54.  (February,  1978) 


BIPHASIC  P ARTICULATE  MEDIA  FOR  THE  CULTURE  OF 

FILTER-FEEDERS  l>  2 

D.  E.  CONKLIN  AND  L.  PROVASOLI 

The  Bodega  Marine  Laboratory,  I'niirrsity  of  California,  Davis,  P.O.  Box  247,  Bodega  Bay, 
California  94923;  and  Haskins  Laboratories,  Biology  Department,  Yale  University, 

AV?v  Haven,  Connecticut  06520 

The  principal  conduit  of  nutrients  between  the  primary  producers  and  higher 
trophic  levels  in  aquatic  ecosystems  is  the  micro-crustaceans.  These  herbivores 
which  feed  in  nature  on  phytoplankton  plus  bacteria  and  fine  detritus  are,  in  turn, 
preyed  on  by  various  carnivores,  such  as  small  fish.  While  the  trophic  role  of  filter- 
feeding  crustaceans  in  aquatic  food  chains  has  been  extensively  documented,  little  is 
known  concerning  their  specific  nutritional  requirements.  One  reason  for  this 
deficiency  has  been  the  lack  of  artificial  media  which  meet  their  specialized  require- 
ments as  phagotrophs. 

A  new  type  of  media  in  which  nutrients  are  supplied  as  both  particles  and  solutes 
(biphasic  participate  media)  has  led  to  the  establishment  of  a  number  of  nutrient 
requirements  for  two  of  these  crustaceans.  The  first  chemically  defined  medium  of 
this  type  allowed  good  survival  and  rapid  growth  from  newborn  to  adult  stages  of 
the  amphigonic  race  of  the  brine  shrimp,  Artemia  salina,  but  the  same  medium  sup- 
ported growth  only  to  juveniles  for  the  parthenogenetic  race  (Provasoli  and 
D'Agostino,  1969).  A  freshwater  version  of  this  medium  for  DapJmia  magna  gave 
similar  results;  growth  to  adult  with  only  occasional  progeny  (Provasoli,  Conklin 
and  D'Agostino,  1970).  While  formulation  of  media  supporting  growth  to  adult 
stages  is  essential  in  defining  cultural  conditions  and  the  major  nutritional  require- 
ments, the  lack  of  fertility  of  the  animals  indicated  that  these  media  were  still 
nutritionally  incomplete. 

The  missing  fertility  factors  in  filter-feeding  crustaceans  were  studied,  using  the 
water  flea  Moina  inacrocopa  auicricana  which  is  viviparous,  parthenogenetic,  and 
has  a  much  shorter  life  cycle.  Molna  was  eventually  grown  for  >200  germ-free 
consecutive  generations  on  three  artificial  media — one  of  which  is  almost  defined. 
This  report  describes  the  compounding  of  nutrient  particles  and  discusses  the  pos- 
sibility of  using  similar  media  to  satisfy  the  phagotrophic  requirements  of  other 
filter-feeding  invertebrates. 

1  It  has  long  been  the  policy  of  THE  BIOLOGICAL  BULLETIN  not  to  accept  methodo- 
logical papers  "which  describe  only  a  new  technique  or  method"  without  extensive  experimental 
results  resulting  from  its  use.  In  view  of  the  difficulties  encountered  in  earlier  attempts  at  the 
axenic  culture  of  filter-feeders,  and  the  importance  of  these  techniques  to  future  studies  in  the 
physiology  and  productivity  of  a  variety  of  aquatic  invertebrates,  it  seemed  appropriate  to  make 
an  exception  in  this  case — Editor. 

-  \York  done  in  partial  fulfillment  of  the  requirements  for  Ph.D.  degree  at  New  York 
University.  Supported  by  XSF  grants  GB-19143  and  GA-33480  (Biological  Oceanography). 

47 


48  D.  K.  CONKLIN  AND  L.  PROVASOLI 

MATERIALS  AND  A!ETIIODS 

The  original  culture  of  M.  inacrocopa  americana  was  obtained  from  Dr.  James 
Murphy  of  the  Rockefeller  University.  Following  his  suggestion  (Murphy,  1970), 
monoxenic  cultures  were  maintained  using  the  algal  species,  Chlauiydonwnas  rcin- 
liardii,  until  an  adequate  artificial  medium  (E  medium)  was  developed.  Early  test- 
ing of  artificial  media  was  complicated  by  the  necessity  of  eliminating  the  algal 
cells.  This  was  done  by  5-10  consecutive  transfers  of  several  animals  in  sterile 
media  containing  starch  particles.  Ingestion  of  the  particles  cleared  the  gut  of  algal 
cells  which  were  eliminated  from  the  medium  by  the  repeated  dilutions.  Once  the 
E  medium  (supplemented  with  lipid-rich  particles  containing  serum,  egg  yolk  and 

TABLE  I 

Artificial  media 

E.  medium:  basal  medium  98  nil  +  2  ml  trigel  particles  +  0.2  nil  egg  particles;  pH 
7.6-7.8. 

FP  medium:  basal  medium  97  ml  +  2  ml  trigel  particles  +  1  ml  FP  particles;  pH 
7.6-7.8. 

F\  medium:  basal  medium  97  ml  +  2  ml  SA  gel  particles  +  1  ml  FV  particles;  pH 
7.6-7.8. 

Particles 

Trigel  particles:  2  nil  supply  15  mg  egg  albumin  +  10  mg  rice  starch  +  5  mg  dry 
beef  serum. 

Egg  particles:  0.2  ml  supply  10  mg  egg  yolk  +  2  mg  vitamin  E  (type  II,  Sigma  Co.) 
+  0.5  mg  calciferol. 

FP  particles:  1  ml  supplies  4.5  mg  albumin  fraction  V  +  3  mg  vitamin  E  (type  II) 
+  1.5  mg  egg  lecithin  +  0.75  mg  calciferol. 

SA  gel  particles:  2  ml  supply  15  mg  egg  albumin  +  10  mg  rice  starch. 

FV  particles:  1  ml  supplies  6  mg  albumin  fraction  V  +  1.5  mg  egg  lecithin  +  1  mg 
BHT  (butylated  hydroxytoluene)  +  1  mg  calciferol  +  0.5  mg  /3-carotene  +  2  mg 
dl-a  tocopherol  +  1  mg  palmitic  acid  +  0.5  mg  oleic  acid  +  1  mg  linoleic  acid  +  1.5  mg 
linolenic  acid. 

Common  basal  medium  (per  cent  w  or  v/v) 

KC1,  3  mg;  MgSO4-7  H,O,  4  mg;  Ca  (as  Cl"),  2  mg;  K3PO4,  2  mg;  Na2SiO»-9  H2O, 
2  mg;  metal  mix  PI  I,  1  ml  (1  ml  contains  Na2EDTA,  1  mg ;  Fe,  0.01  mg ;  B,  0.2  mg; 
Mn,  0.04  mg;  Zn,  0.005  mg ;  Co,  0.001  mg) ;  Fe  (as  (NH4)2  H  citrate),  0.05  mg;glycyl- 
glycine,  50  mg,  pH  8.0  [TRIS  buffer  (Sigma  Co.)  is  toxic  for  Artemia,  Daphnia  and 
Moina  at  50  mg%.  TES  buffer  (Sigma  Co.)  is  nontoxic  at  100  mg%  for  Molna~];  nucleic 
acid  mix  V,  2  ml  (1  ml  contains  adenylic  acid,  20  mg ;  guanylic  acid,  10  mg ;  cytidylic 
acid,  10  mg;  thymidine,  10  mg;  dissolve  in  alkali,  adjust  to  pH  8.0);  DF  2,  1  ml  (1  ml 
contains  Tween  60,  2  mg;  Tween  80,  2  mg;  rutin,  0.5  mg ;  oxbile  extract  (Nutritional 
Biochem  Co.),  1  mg;  disperse  and  emulsify  components;  adjust  to  pH  8.0) ;  Cholesterol, 
0.6  mg  (dissolved  in  95%  ethanol,  squirted  into  boiling  water,  ethanol  boiled  off;  forms 
fine  crystalline  precipitate);  amino  acids  mix  III,  1  ml  (1  ml  contains  L-isoleucine, 
10  mg;  L-lysine  HC1,  L-glutamic  acid,  L-histidine  base,  L-threonine,  L-methionine, 
L-leucine,  L-valine,  L-proline,  1  mg  each;  L-arginine  base,  L-tyrosine,  L-serine,  glycine, 
L-tryptophane,  0.5  mg  each);  vitamin  mix  M1B,  1  ml  (1  ml  contains  thiamine  HC1, 
0.5  mg;  nicotinamide,  1.5  mg;  pyridoxine  HC1,  0.2  mg;  biotin,  0.06  mg;  putrescine-2  HCI, 
0.1  mg;  Vitamin  Bi2,  0.002  mg;  choline  H2  citrate,  0.2  mg;  riboHavin,  0.2  mg;  folic  acid, 
0.1  mg;  Ca  pantothenate,  4  mg) ;  liver  infusion  L  25  (Oxoid,  Flow  Labs,  Rockville,  Md.), 
70  mg  (does  not  dissolve  completely;  upon  autoclaving  in  medium  forms  a  brown  pre- 
cipitate essential  for  growth).  Adjust  pH  of  basal  medium  to  pH  7.6-7.8. 


PARTICULATE  MEDIA  FOR  FILTER-FEEDERS  49 

vitamins  Do  and  E;  Table  I)  was  developed,  it  was  used  both  as  the  maintenance 
medium  and  the  control  medium  during  further  work  on  substitution  of  serum 
and  egg  yolk  with  more  chemically  denned  particles.  Transfer  techniques  used  for 
the  bacteria-free  Moina  studies  were  essentially  those  developed  for  Artemia 
nauplii  (Provasoli,  Shiraishi  and  Lance,  1959). 

The  general  form  of  the  media  is  presented  in  Table  I.  This  type  of  biphasic 
media  is  similar  to  those  developed  for  culturing  Artemia.  The  liquid  phase  con- 
tains salts  and  trace  metals,  pH  and  metal  buffers,  ami  no  acids,  nucleic  acids  and  a 
mixture  of  water-soluble  vitamins.  The  solid  phase  is  a  slurry  of  fine  (up  to  30  ju,m) 
particles  of  proteins,  carbohydrates,  and  lipids.  The  addition  of  lipid-rich  particles 
proved  necessary  for  continuous  generations  of  Moina. 

Particle  preparation 

SA  gel.  Dissolve  completely  750  mg  of  egg  albumin  (2X  cryst.,  Sigma  Chem. 
Co.)  in  30  ml  H2O  before  adding  500  mg  of  rice  starch.  The  mixture  is  then 
homogenized  in  a  Yirtis  homogenizer  model  "23"  (container  #16-117)  for  a  few- 
minutes  using  two  straight  blades  at  right  angles  (Virtis  blade  #2-16-108).  The 
mixture  is  autoclaved  (20  min  at  20  Ib),  cooled,  homogenized  for  another  5  min  at 
medium  to  high  speed  and  reautoclaved.  Following  a  final  homogenization,  the 
suspension  is  diluted  to  100  ml  with  HoO  resulting  in  a  fine,  milky-white  liquid. 
Autoclaving  the  gel  twice  prevents  reaggregation  of  the  particles  during  storage 
and  also  during  the  autoclaving  of  the  complete  medium. 

Trigcl.  Water  is  added  drop-wise  to  250  mg  of  dried  beef  serum,  avoiding 
lumps  which  would  stick  to  the  container  wall,  until  the  serum  is  completely  dis- 
solved. The  mixture  is  then  brought  to  30  ml  with  HoO.  Then  750  mg  albumin 
and  500  mg  rice  starch  are  added,  and  the  mixture  is  homogenized  and  autoclaved 
following  the  procedure  detailed  for  the  SA  gel.  The  final  appearance  of  the  trigel 
is  a  fine  light-brown  suspension. 

Egg  particles.  A  fresh  egg  yolk,  free  of  albumin,  is  transferred  without  break- 
ing to  a  30  ml  beaker.  The  membrane  is  penetrated  with  a  5  ml  pipette,  and  the 
yolk  material  sucked  up.  Three  ml  of  yolk  is  allowed  to  flow  from  the  pipette  into 
a  test  tube  (16  X  75  mm).  Free  flow  insures  more  repeatability  than  blowing  out 
since  it  avoids  differing  amounts  of  yolk  coating  the  pipette  wall.  Add  150  mg  of 
ergocalciferol  to  0.6  ml  of  a  tocopherol  concentrate  (a-tocopherol  type  II,  Sigma 
Chem.  Co.)  and  triturate  with  a  glass  rod  until  completely  dissolved.  After  addi- 
tion of  10  ml  of  HoO,  the  mixture  is  emulsified  on  the  "Vortex  Genie"  (Scientific 
Industries  Inc.,  Queens  Village,  New  York)  at  top  speed.  The  mixture  is  trans- 
ferred into  a  Virtis  container  16-1 17.  rinsing  the  test  tube  twice  with  10  ml  of  H2O 
each  time  and  emulsified  further  with  3  min  of  homogenization  with  the  double 
blades.  The  emulsion  is  heated  in  a  water  bath  on  a  hot  plate  with  constant  stirring 
until  coagulated  in  large  floes.  The  egg  mixture  is  then  put  through  two  cycles  of 
autoclaving,  cooling,  addition  of  5  ml  HL»O,  and  homogenization  for  3  min.  Finally, 
the  mixture  is  diluted  to  60  ml  with  HoO.  The  resulting  light  yellow  suspension  is 
stored  in  a  glass-stoppered  bottle,  flushed  with  Xo,  and  refrigerated.  Even  though 
autoclaved  twice,  the  egg  particles  tend  to  aggregate  on  storage  and  must  be 
thoroughly  agitated  before  use. 


50  D.  E.  CONKLIN  AND  L.  PROVASOLI 

FV  particles.  A  more  defined  mixture  of  lipids  was  specifically  tailored  to  the 
needs  of  Moina  and  replaced  the  serum  and  egg  yolk  supplements  of  the  E  medium. 
To  compensate  for  the  emulsifying  properties  of  the  egg  yolk,  egg  lecithin  is  used. 
Add  75  mg  of  egg  lecithin  and  300  mg  of  albumin  (Fraction  V,  Sigma  Chem.  Co.) 
to  25  ml  of  H2O  in  a  Yirtis  flask  (16-115).  This  flask  has  an  enlarged  bottom  with 
small  fluting  and  a  side-arm  capped  with  a  small  rubber  plug  on  the  top  of  the 
enlarged  bottom.  The  lipid  solution  is  prepared  separately  in  a  short  test  tube. 
The  dry  solids  are  added  first,  in  the  following  order :  butylated  hydroxytoluene 
(BHT),  100  mg;  ergocalciferol,  100  mg;  /3-carotene,  30  mg;  and  palmitic  acid,  100 
mg.  Then  in  order:  dl-a-tocopherol.  0.2  ml;  linolenic  acid,  0.15  ml;  linoleic  acid, 
0.1  ml;  oleic  acid,  0.05  ml;  and  1.5  ml  of  acetone.  Stirring  with  a  glass  rod  and 
use  of  the  "Vortex  Genie"  helps  to  dissolve  the  mixture  completely.  One  ml  of  the 
lipid  mixture  is  drawn  into  a  small  hypodermic  syringe  with  a  thin  needle.  The 
albumin  and  lecithin  are  homogenized  thoroughly  for  2-3  min  at  top  speed  (with  the 
2  straight  blades)  before  the  1  ml  of  lipid  mixture  is  slowly  squirted  into  the  Virtis 
container  through  the  rubber  cap  covering  the  side-arm.  Homogenization  is  con- 
tinued for  8  min  at  top  speed,  followed  by  autoclaving  and  cooling.  The  appear- 
ance after  autoclaving  is  not  uniform  :  a  thin  skin  of  coagulated  material  overlaps 
the  liquid  containing  a  flocculent  mass.  The  skin  and  the  coagulum  are  mixed  and 
resuspended  with  a  glass  rod,  then  homogenized  for  5  min.  The  above  procedure 
of  autoclaving  and  homogenization  for  5  min  is  repeated  once  more  and  the  final 
volume  brought  to  100  ml.  The  final  appearance  is  a  brownish-red  suspension  of 
fine  particles. 

Carotene  is  difficult  to  dissolve  and  is  replaceable  with  0.03  ml  retinol  palmitate 
(Type  IV,  Sigma  Chem.  Co)  resulting  in  a  more  homogeneous  initial  coagulum. 
Increasing  the  fat-binding  albumin  fraction  V  (>600  mg)  inhibits  the  growth  of 
Moina.  However,  we  found  recently  that  when  egg  albumin,  which  can  be  used  in 
higher  concentrations,  is  substituted  for  fraction  V,  the  resulting  particles  are  again 
more  homogeneous.  Initially  300  mg  of  egg  albumin  plus  75  mg  of  lecithin  are 
homogenized  together  in  25  ml  H2O.  After  adding  the  fat  solution  to  the  mixture, 
it  is  homogenized  for  5  min  at  top  speed,  then  an  additional  500  mg  of  egg  albumin 
is  added ;  followed  with  another  5  min  homogenization.  The  emulsion  from  the 
Virtis  container  is  transferred  to  a  600  ml  beaker  and  coagulated  in  a  boiling  water 
bath  with  continuous  stirring.  Following  this  rapid  coagulation,  the  lipid  particle 
mixture  is  autoclaved  and  homogenized  twice  as  outlined  above  and  brought  to  100 
ml.  All  the  lipid  particle  mixes  are  stored  refrigerated  in  glass  stoppered  bottles 
which  have  been  flushed  with  N2. 

FP  particles.  Another  lipid  particle  was  also  successful  in  replacing  the  egg 
particle  (medium  FP,  Table  I,  Conklin  and  Provasoli,  1977).  Dissolve  225  mg  of 
albumin  fraction  V  in  20  ml  H2O,  add  75  mg  of  egg  lecithin  in  a  Virtis  flask  with 
side  arm ;  homogenize  for  3  min  at  top  speed.  Then  add,  as  above,  1  ml  lipid  mix- 
ture [0.15  ml  a-tocopherol  type  II  (Sigma  Chem.  Co.)  +  37.5  mg  ergocalciferol 
dissolved  in  1  ml  acetone]  through  the  side  arm  and  homogenize  for  8  additional 
minutes ;  follow  as  for  FV  particles  with  2  cycles  of  autoclaving,  cooling,  homogeniz- 
ing and  bring  to  50  ml.  The  simpler  FP  medium  may  be  useful  for  other  filter 
feeders. 


PARTICULATE  MEDIA  FOR  FILTER-FEEDERS  51 

"While  the  proportions  in  the  medium  of  the  SA  gel,  trigel.  and  the  FY  and  FP 
particles  may  be  varied  to  suit  other  filter  feeders,  modifications  in  the  composition 
of  the  gels  and  lipid  particles  should  not  exceed  the  limited  binding  power  of  the 
albumin.  To  insure  a  good  coagulation  and  protein  binding  and  to  avoid  separa- 
tion of  the  lipids,  it  is  necessary  to  use  a  small  amount  of  H2O  (20-30  ml)  in  the 
initial  mixture  that  is  homogenized  and  coagulated  for  the  first  time  by  heat  or 
autoclaving.  The  particles  thus  produced  are  stabilized  by  the  second  autoclaving 
and  after  the  final  homogenization  can  be  dispersed  in  a  large  volume  of  H2O 
(50-100  ml  or  more)  without  changing  their  physical  properties. 

RESULTS 

The  media  are  biphasic.  The  mineral  base  [minerals,  trace  metal  mix,  and 
glycylglycine  (at  pH  7.8)]  was  a  modification  of  the  medium  formulated  for 
Daphnia  magna  (D'Agostino  and  Provasoli,  1970)  which  proved  satisfactory  for 
rearing  this  cladoceran  in  dixenic  culture  on  Chlamydomonas  rcinhardii  and 
Scenedesmus  obliquus. 

Assuming  that  essential  nutrients  for  Art  curia  might  also  be  required  by  Moina, 
various  combinations  of  amino  acid,  nucleic  acid  and  vitamin  mixtures  were  used, 
and  various  quantities  and  ratios  of  starch  and  protein  were  co-gelled  into  fine 
homogeneous  particles.  A  striking  difference  was  seen  in  protein :  starch  ratios. 
Specimens  of  Moina,  as  well  as  those  of  Daphnia,  seem  to  prefer  a  more  even  ratio 
of  protein:  starch  (P:  S  =:  1  :  0.5-2.0)  in  contrast  to  Artcuria  which  needs  a  high 
starch  ratio  (P:S=:  1:5).  On  the  contrary,  the  requirements  for  most  water 
soluble  nutrients  were  similar  although  adjustments  in  concentrations  were  neces- 
sary. Media  at  this  stage  did  not  support  consecutive  generations  for  Moina 
macrocopa.  Some  adults  were  produced  but  the  sparse  progeny  did  not  reach 
adulthood. 

Failure  of  satisfactory  viability  presumably  was  due  to  lipid  deficiencies :  many 
insects  need  several  fatty  acids  and  some  require  tocopherol  for  fecundity  and  all 
need  sterols  (Dadd,  1973).  M.  rcctirostris  produced  males,  females  and  ephippial 
eggs  in  bacterized  cultures  fed  defatted  yeast  supplemented  with  olive  oil  and 
ergosterol  (von  Dehn,  1955)  ;  fertility  of  Daphnia  magna  under  similar  conditions 
was  thought  to  be  restored  by  vitamin  E  (Viehoever  and  Cohen,  1938). 

Early  attempts  to  supply  lipids  as  emulsions  did  not  prove  very  useful.  Efforts 
were  then  directed  toward  producing  lipid-rich  solid  particles.  A  particle  made  up 
of  starch,  protein  and  serum  (trigel;  Table  I)  permitted  one  or  two  more  gen- 
erations. Additions  of  ergocalciferol  and  the  vitamin  E  concentrate  in  an  egg  yolk 
carrier  resulted  in  a  repeatable  preparation  of  highly  nutritious  particles.  The 
Sigma  Chemical  Co.  "Type  II"  a-tocopherol,  an  equal-part  mixture  of  a-tocopherol 
and  a  vegetable  oil,  supplied  a  convenient  array  of  fatty  acids  and  an  antioxidant. 
Coagulated  egg  yolk,  added  primarily  as  a  carrier  for  the  vitamin  E  oil,  presumably 
also  supplied  a  number  of  nutrients ;  however,  the  egg  yolk  alone  was  poor  or 
inhibitory.  In  this  lipid-rich  medium,  Daphnia  magna  produced  5  or  6  successive 
parthenogenetic  generations,  while  M.  macrocopa  continued  to  reproduce  without 
decline  in  fertilitv. 

•" 

A  suitable  lipid-rich  particle   (FY)  was  eventually  devised  with  albumin  frac- 


52  D.  E.  CONKLIN  AND  L.   PK<  )V.\SOLI 

tion  V  as  the  fat-acceptor  and  coagulant.  This  particle  served  to  define  the  need 
of  Moina  for  fatty  acids,  ergocalciferol  and  a-tocopherol.  Details  on  nutritional 
requirements  are  given  elsewhere  (Conklin  and  Provasoli,  1977)  ;  it  suffices  to  say 
that  Moina  also  needs  intact  nucleic  acids  and  water  soluble  vitamins  and  that  the 
concentrations  given  for  the  Fl  medium  are  close  to  optimal  under  our  conditions 
(22-24°  C,  subdued  light).  All  the  solids  used  in  biphasic  media  are  a  slurry  of 
particles  ranging  up  to  30  /xm  in  diameter.  When  added  to  the  media,  the  particles 
remain  in  suspension  for  several  hours.  For  maximum  efficiency  of  ingestion  the 
particles  are  resuspended  twice  a  day  by  shaking  the  test  tubes  on  a  "Vortex  Genie" ; 
the  animals  are  not  harmed  by  the  vigorous  mixing. 

Media  E  and  Fl  allow  a  generation  time  of  4-6  days,  clutches  of  4-8  newborn, 
without  decline  in  fertility  for  over  200  consecutive  parthenogenetic  generations. 
The  effectiveness  of  each  variable  in  the  diet  was  gauged  from  the  number  of  ani- 
mals produced  from  a  single  female  in  a  week,  day  1  being  the  day  when  the  female 
produced  the  first  brood.  The  variability  due  to  the  time  needed  for  the  inoculated 
animal  (newborn  or  young)  to  produce  offspring  was  thus  avoided.  In  10  ml  of 
complete  medium  at  the  end  of  day  7,  the  Moina  population  comprised  three  gen- 
erations :  the  original  female,  females  of  the  first  and  second  clutch,  and  their  com- 
bined progeny,  i.e.,  about  13  adults  and  close  to  100  newborn.  Growth  and  repro- 
duction ceased  in  about  two  weeks  when  almost  all  the  particles  in  the  10  ml  of 
medium  were  ingested. 

DISCUSSION 

This  report  on  techniques  for  producing  several  kinds  of  nutrient  particles  is 
motivated  by  the  hope  that  other  researchers  may  formulate  better  participate 
media  and  adapt  this  type  of  media  to  satisfy  the  particulate  requirements  of  other 
filter-feeders.  Protozoa,  sponges,  rotifers,  molluscs  and  many  Crustacea  are  filter- 
feeders  throughout  life  or  at  least  in  the  early  larval  stages ;  some  primitive  chor- 
dates  such  as  sea  squirts  and  salps,  and  some  fishes  are  also  filter-feeders. 

Our  experience  with  Artcinia  and  Moina  and  the  work  of  Akov  (1962)  and 
Dadd  (1972)  on  mosquito  larvae,  indicate  that  success  in  growing  filter-feeders  de- 
pends on  two  equally  important  factors :  supplying  all  the  essential  nutrients  for 
growth  and  reproduction,  and  compounding  the  media  so  that  the  nutrients  are  ac- 
ceptable and  readily  available  to  the  animals.  The  biphasic  media  for  Moina 
satisfy  both  requisites  and  result  in  rapid  growth,  high  fertility,  and  continuous 
parthenogenetic  reproduction. 

The  compromise  found  experimentally  effective  was  that  the  nutrients  which  are 
required  in  large  amounts  by  crustaceans  for  rapid  growth  must  be  supplied  as  fine 
particles  (e.g.,  the  amino  acids  as  precipitated  proteins  and  the  energy  sources  as 
insoluble  starch  and/or  fats).  The  soluble  nutrients  were  added  at  noninhibitory 
concentrations  and  high  enough  to  compensate  for  the  poor  uptake  of  solutes  by 
crustaceans.  Uptake  through  the  thick  chitinous  exoskeleton,  except  for  the  areas 
used  for  osmoregulation,  is  apparently  minimal;  most  of  the  uptake  is  through 
imbibition  of  water  while  ingesting  the  bolus  and  perhaps  through  anal  uptake 
(Fryer,  1970).  Stephens  and  Schinske  (1961)  found  that  of  the  11  phyla  tested, 
only  the  6  crustaceans  tried  were  unable  to  take  up  labeled  amino  acids  from  the 


PARTICULATE  MEDIA    FOR   FILTER-FEEDERS  53 

environmental  water.  Some  uptake  of  palmitate  and  glucose  (at  5-250  /u.Ci, 
respectively)  was  shown  by  Sargent  and  Lee  (1975),  but  evidently  this  uptake  is 
not  sufficient  to  support  the  nutritional  requirements.  We  found  that  replacements 
of  starch  and  protein  particles  with  soluble  carbohydrates  and  ammo  acid  mixtures 
was  partial  and  inefficient:  growth  rates  were  slowed  2-3 X  and  the  solutes  had 
1/20-1/60  X  the  efficacy  of  participates  for  Arteinia.  Therefore,  crustaceans  may 
be  considered  as  obligate  phagotrophs. 

On  the  contrary,  the  work  of  Stephens  and  Schinske  (1961)  and  later  work  of 
Stephens  (1975)  and  Wright  and  Stephens  (1977)  shows  that  soft-bodied  marine 
invertebrates  are  able  to  take  up  and  incorporate  considerable  amounts  of  dissolved 
carbon  sources  and  ami  no  acids  at  the  very  low  concentrations  present  in  sea  water. 
Hence,  the  organic  solutes  in  biphasic  media  could  be  taken  up  by  the  soft-bodied 
invertebrates,  and  if  the  rate  of  uptake  is  considerable  it  might  be  necessary  to  lower 
the  concentration  of  the  present  media  to  compensate  for  the  increased  uptake  of 
solutes.  Yet,  even  for  these  "permeable"  filter-feeders  the  need  for  participates 
(phagotrophy)  may  be  postulated,  because  filter-feeding  is  a  very  effective  gathering 
process  as  indicated  by  the  nutritional  efficiency  of  particles  over  solutes. 

While  most  filter-feeders  living  in  oceanic  (even  coastal)  waters  depend  mostly 
on  phagotrophy,  such  an  assumption  mav  not  be  valid  for  environments  high  in 
soluble  organic  matter  (i.e.,  high  domestic  pollution,  where  death  and  decay  of  ani- 
mals or  plant  blooms  occur,  and  perhaps  in  aerobic  detrital  sediments).  There, 
organisms  utilizing  phagotrophy  as  well  as  osmotrophy  (effective  uptake  of  solutes) 
would  have  a  great  advantage :  they  would  not  depend  solely  on  the  transforma- 
tion of  solutes  into  bacteria,  being  able  to  take  up  solutes  directly.  Rasmussen 
(1976)  has  brilliantly  demonstrated  that  the  freshwater  ciliate  Tetrahymena  pyri- 
fonnis  is  almost  equally  efficient  as  an  osmotroph  or  a  phagotroph  and  that  the 
uptake  of  solutes  by  Tetrahymena  is  almost  as  efficient  as  bacterial  uptake.  Perhaps 
filter-feeders  in  organically-rich  environments  employ,  in  different  degrees,  the  same 
strategy. 

Biphasic  media  used  with  germ-free  techniques  could  be  a  useful  tool  in  rearing 
a  variety  of  filter-feeders  and  in  defining  their  nutritional  requirements.  They  offer 
a  new  approach  because  of  the  great  experimental  versatility  of  preparing  particles 
of  different  composition  and  ratios  and  of  the  possibility  of  supplying  lipids  more 
efficiently  than  with  emulsions  which  are  often  inhibitory. 

Once  nutrient  requirements  are  understood,  it  may  be  possible  to  remove  the 
limitation  of  germ-free  handling  by  using  microencapsulation  techniques  (Jones, 
Munford,  and  Gabbott,  1974).  Further  improvements  of  biphasic  media  under 
both  bacterized  and  bacteria-free  conditions  should  lead  to  the  definition  of  the 
nutritional  requirements  of  ecologically  and  commercially  important  filter-feeders. 
Hopefully,  this  data  can  also  be  applied  in  the  formation  of  efficient  diets  for  a 
number  of  anticipated  aquaculture  species  (Provasoli,  1976) . 

SUMMARY 

1.  Over  200  parthenogenetic  generations  of  the  freshwater  Cladocera  Moina 
inacrocopa  were  obtained  aseptically  in  three  artificial  media. 

2.  The  media  have  two  phases:    a  liquid  phase  supplying  mineral  salts,  water 


54  D.  K.  CONKLIN  AND  L.  PROVASOLI 

soluble  vitamins,  nucleic  acids,  and  a  liver  extract  and  a  fine  particulate  phase  made 
from  coagulated  proteins,  starch  and  lipid  factors. 

3.  The  particulate  phase  supplies  the  bulk  nutrients  very  efficiently;  hence,  this 
type  of  media  may  be  useful  for  growing  other  filter-feeders. 

LITERATURE  CITED 

AKOV,  S.,  1962.     A  qualitative  and  quantitative  study  of  the  nutritional  requirements  of  Aeties 

acgypti  larvae.    /.  Insect  Pliysiol.,  8  :  319-335. 
CONKLIN,  D.  E.,  1973.     Nutritional  requirements  of  Moina  niacrocopa  in  axenic  culture.    Ph.I>. 

Thesis,  New  York  University,  New  York,  90  pp.   (Diss.  Abstr.,  34-03:  989-B ;  order 

no.  73-19,  911.) 
CONKLIN,   D.  E.,  AND  L.   PROVASOLI,   1977.   Nutritional   requirements  of  the  water   flea  Moina 

macrocopa.   Biol.  Bull,  152:  337-350. 
DADD,   R.   H.,   1972.     Ambiguities  in  the  interpretation  of  growth  experiments  with   mosquito 

larvae  in  semi-synthetic  dietary  media.     Pages  199-209  in  J.  Rodriguez,  Ed.,  Insect  and 

mite  nturitinn,  North-Holland  Publishers  Co..  Amsterdam. 
DADD,  R.  H.,   1973.     Insect  nutrition:    current  developments  and  metabolic   implications.  Ann. 

Rev.  Entomol,  18:  381-420. 
D'AGOSTINO,  A.  S.,  AND  L.  PROVASOLI,  1970.  Dixenic  culture  of  Daphnia  inagna  Straus.  Biol. 

Bull,  139 :  485-494. 
DEHN,  M.  VON,  1955.  Die  geschlechtsbestimmung  der  daphniden.    Die  bedeutung  der  fettstoffe 

untersucht  an  Mania  rectirostris.  Zool.  Jb.  Abt.  Allg.  Zool.  Physiol.  Tiere.,  65:  334-356. 
FRYER,  G.,  1970.     Defaecation  in  some  macrothricid  and  chydorid  cladocerans,  and  some  prob- 
lems of  water  intake  and  digestion  in  the  Anomopoda.    Zool.  J.  Linn.  Soc.,  49:  255-269. 
JONES,  D.  A.,  J.  G.  MUNFORD,  AND  P.  A.  GABBOTT,  1974.     Microcapsules  as  artificial  particles 

for  aquatic  filter-feeders.    Nature,  247  :  233-235. 
MURPHY,  J.  S.,  1970.    A  general  method  for  the  monoxenic  cultivation  of  the  Daphnidae.  Biol. 

Bull.,  139:  321-332. 
PROVASOLI,    L.,    1976.      Nutritional    aspects    of   crustacean    aquaculture.      Pages    13-21    in    K.S. 

Price,  W.  N.  Shaw  and  K.  S.  Danberg,  Eds.,  Proceedings  of  the  First  International 

Conference   on   Aquaculture   Nutrition.      College    of    Marine    Sciences,    University    of 

Delaware,  Newark,  Delaware. 
PROVASOLI,   L.,   AND   A.   S.   D'AGOSTINO,    1969.     Development   of   artificial   media   for   Artcinia 

salina.  Biol.  Bull.,  136  :  434-453. 
PROVASOLI,   L.,   D.   E.   CONKLIN,  AND  A.   S.   D'AGOSTINO,   1970.     Factors   inducing   fertility   in 

aseptic  Crustacea.     Hclgol.  U'iss.  Mceresuntcr.,  20:  443-454. 
PROVASOLI,  L.,  K.  SHIRAISHI,  AND  J.  R.  LANCE,  1959.     Nutritional  iodiosyncrasies  of  Artcinia 

and  Tigriopus  in  monoxenic  culture.     Ann.  X.  Y.  Acad.  Sci.,  77:  250-261. 
RASMUSSEN,  L.,  1976.     Nutrient  uptake  in   Tctrah\incna  p\rijorinix.   Carlsbcrg  Res.   Coinmun., 

41 :  143-167. 
SARGENT,  J.  R.,  AND  R.  F.  LEE,  1975.     Biosynthesis  of  lipids  in  zooplankton  from  Saanich  Inlet, 

British  Columbia,  Canada.     Mar.  Biol.,  31:  15-23. 
STEPHENS,   G.   C.,   1975.     Uptake  of  naturally  occurring  primary  amines   by  marine   annelids. 

Biol.  Bull.,  149:  397-407. 
STEPHENS,  G.  C.,  AND  R.  A.  SCHINSKE,  1961.     Uptake  of  amino  acids  by  marine  invertebrates. 

Liinnol.  Occanogr.,  6:  175-181. 
VIEHOVER,  A.,  AND  J.   COHEN,   1938.     The  response  of  Daphnia   magna  to  vitamin  E.  Am.  J. 

Pharm.,  110:  297-315. 
WRIGHT,   S.  H.,  AND  G.  C.  STEPHENS,   1977.     Characteristics  of  influx  and  net  flux  of  amino 

acids  in  Mytilus  ealifoniiunus.  Biol.  Bull.,  152:  295-310. 


Reference:  Biol.  Bull.,  154:  55-67.  (February,  1978) 


DEVELOPMENT  OF  THE  DIMORPHIC  CLAW  CLOSER  MUSCLES  OF 

THE  LOBSTER  HOMARUS  AMERICANUS.    III.  TRANSFORMATION 

TO  DIMORPHIC  MUSCLES  IN  JUVENILES 

C.  K.  GOVIXD  AND  FRED  LANG 

Scarborough  Collcyc,  University  of  Toronto,  H'cst  Hill,  Ontario,  Canada  MIC  lA-f;  and 

Boston  University  Marine  Proyrani,  Marine  Biological  Laboratory, 

Woods  Hole  Massachusetts  02543 

The  asymmetry  observed  in  the  chelipeds  of  many  crustaceans  presents  inter- 
esting problems  in  a  number  of  areas,  including  development,  behavior,  and  neuro- 
muscular  physiology.  While  asymmetry  is  fixed  in  some  animals  (Przibram,  1931), 
there  are  a  number  of  examples  where  it  has  been  demonstrated  that  claw  type  can 
be  "reversed."  That  is,  loss  of  one  claw,  usually  the  larger  or  "crusher"  will  result 
in  transformation  of  the  remaining  smaller  claw,  the  "cutter,"  into  a  crusher.  The 
regenerated  claw  will  then  become  a  cutter  (Przibram,  1931 ;  Hamilton,  Nishimoto, 
and  Halusky,  1976).  Furthermore,  in  some  of  the  species  in  which  reversal  has 
been  demonstrated,  it  has  also  been  shown  that  the  claws  are  used  for  different 
behaviors  (e.g.,  Alplicns,  Przibram,  1931;  Calappa,  Shoup,  1968;  Lewis,  1969; 
Callincctes,  Hamilton  ct  a!.,  1976).  Thus,  it  would  be  of  interest  to  ascertain  the 
mechanisms  underlying  both  the  morphogenetic  changes  which  are  manifested  and 
also  the  possible  central  nervous  system  modifications  which  in  some  cases  must 
also  occur.  There  presently  is  little  information  regarding  the  neuromuscular 
physiology  or  development  of  any  of  the  aforementioned  crustaceans,  either  before 
or  after  reversal.  Ritzman  (1974)  has  described  the  neural  mechanisms  under- 
lying closure  of  the  large  snapping  claw  in  two  species  of  Alphens,  but  has  not  re- 
ported similar  studies  for  the  smaller  "pinch-claw"  or  after  claw  reversal.  While 
it  is  not  yet  known  whether  the  muscle  liber  populations  differ  between  the  two 
claws,  they  are  certainly  used  differently  in  the  behavioral  repertoires  of  the  animals 
(Darby,  1934).  Thus  the  pinch-claw  is  not  merely  a  miniature  snapping  claw 
which  hypertrophies  upon  loss  of  the  snapping  claw. 

Warner  and  Jones  (1976)  have  studied  muscle  liber  properties  in  the  dimorphic 
claws  of  Macropipns  dcpnrator.  Although  the  stouter  chela  of  this  animal  has  a 
higher  mechanical  advantage  than  the  smaller  chela,  there  was  no  consistent  differ- 
ence between  the  muscle  fiber  populations  found  in  each  claw ;  both  claws  contained 
"slow"  type  fibers  with  sarcomere  lengths  of  6-10  /mi. 

In  adult  lobsters  (Hoiiiants  anicricamis)  the  dimorphic  claws  contain  closer 
muscles  which  have  different  populations  of  muscle  fiber  types  (Jahromi  and  At- 
wood,  1971;  Goudey  and  Lang,  1974).  The  fast  acting  cutter  claw  closer  muscle 
is  composed  of  over  60%  short  sarcomere  (2-4  /xm"),  fast  fibers  with  the  remainder 
being  long  sarcomere  (6-12  /mO,  slow  filters.  The  slow  acting  crusher  closer 
muscle  has  virtually  all  long  sarcomere  (>6  p.m],  slow  fibers.  Furthermore,  in  the 
cutter  muscle,  fast  and  slow  fibers  are  regionally  distributed  on  the  inner  aspect 
with  fast  fibers  in  the  dorsal  and  central  medial  sections  and  slow  fibers  in  the 
ventral  sections  (Lang,  Costello,  and  Govind,  1977). 

55 


56  C.  K.  GOVIND  AND  F.  LANG 

In  larval  lobsters  the  claws  arc  identical  and,  indeed,  the  paired  closer  muscles 
have  not  differentiated  into  cutter  and  crusher  types.  In  fact,  the  muscles  are  sym- 
metrical in  early  larval  animals  (stages  1-2),  being  composed  of  ZQ-4Q%  short 
sarcomere,  over  50c/c  intermediate  and  only  \Q%  long  sarcomere  fibers.  In  the  late 
larval  lobsters  (stage  3)  there  is  a  nearly  equal  distribution  of  short,  intermediate 
and  long  sarcomere  fibers  (Lang,  Govincl  and  She,  1977).  Thus,  the  transforma- 
tion of  the  paired  symmetrical  muscles  into  cutter  and  crusher  types  must  occur  in 
postlarval  (juvenile)  forms. 

In  the  larval  stages  (1-3)  and  early  juvenile  stages  (4  and  5)  the  two  claws 
are  identical  in  external  appearance,  both  being  cutter-like  (Herrick,  1896,  1911). 
A  distinct  change  in  external  morphology  of  the  paired  claws  into  cutter  and  crusher 
claws  is  seen  only  at  stage  7  or  8  when  the  cutter  has  a  longer,  more  slender 
shape  and  the  crusher  has  a  larger  blunt  tooth  (Herrick,  1896).  It  is  reasonable 
to  assume  that  the  transformation  in  external  character  signals  muscle  fiber  dif- 
ferentiation in  the  closer  muscle.  Muscle  fiber  types  and  their  distribution  in 
several  juvenile  stages  were  examined  in  this  study  and,  while  differentiation  of  the 
cutter  muscle  occurs  as  early  as  stage  6,  that  of  the  crusher  is  not  usually  completed 
until  at  least  stage  13. 

MATERIALS  AND  METHODS 

Newly  hatched  larval  lobsters  were  obtained  from  the  Massachusetts  State 
Hatchery  on  Martha's  Vineyard  and  reared  in  running  sea  water  tanks  at  20-23°  C 
according  to  the  methods  of  Hughes,  Shleser  and  Tchobanoglous  (1974).  Their 
early  development  consists  of  three  pelagic  mysis  (larval)  stages.  When  they 
molt  to  the  fourth  stage  they  approximate  their  adult  form,  and  during  this  stage, 
or  the  following  one,  they  assume  a  benthic  existence  (Herrick,  1896).  From  the 
fourth  stage  onward,  the  juvenile  lobsters  were  reared  in  individual  trays  (Lang, 
1975)  and  their  growth  followed  for  periods  up  to  two  years. 

Several  animals  were  examined  in  the  early  or  late  period  of  the  molting  cycle. 
In  the  former  case,  animals  were  used  within  one  or  two  days  after  a  molt.  In  the 
latter  case,  two  criteria  were  used  to  establish  that  lobsters  were  in  the  late  part  of 
the  stage,  i.e.,  about  to  molt :  first,  when  molting  had  occurred  in  animals  that  had 
simultaneously  entered  the  same  stage  and  had  been  kept  under  similar  conditions ; 
and  secondly,  the  typical  premolting  behavior  of  failing  to  eat  food  put  into  the  tray. 

The  claw  closer  muscles  were  fixed  with  Benin's  solution  while  the  dactyl  was  in 
the  fully  open  position.  Methods  for  isolating  muscle  fibers  and  measuring  sarco- 
mere lengths  have  been  previously  described  (Lang,  Costello  and  Govind,  1977). 
The  average  sarcomere  length  for  a  fiber  was  established  by  measuring  five  con- 
secutive sarcomeres  in  three  separate  myofibril  bundles.  Sarcomeres  were  sampled 
from  the  inner  aspect  of  the  closer  muscle  which  was  subdivided  into  nine  sections. 
This  partitioned  the  muscle  laterally  into  dorsal,  medial  and  ventral  sections,  and 
transversely  into  proximal,  central  and  distal  sections  (Lang,  Costello,  and  Govind, 
1977).  For  some  stage  4  animals  the  muscle  was  divided  into  only  six  sections  by 
omitting  a  medial  section  and  retaining  only  the  dorsal  and  ventral  sections.  In 
most  animals,  ten  fibers  were  inspected  in  each  section  giving  a  total  sample  of  90 
fibers  for  each  muscle;  in  four  stage  4  animals,  only  60  fibers  were  sampled  from 


CLOSER  MUSCLE  IN  JUVENILE  LOBSTERS 


57 


the  six  sections  for  each  muscle.  It  is  estimated  that  each  closer  muscle  contains 
600-700  fibers;  thus  we  are  sampling  approximately  13-15%  of  the  total  popula- 
tion. However,  the  extremely  small  size  of  the  closer  muscle  may  well  have  intro- 
duced significant  errors  in  the  sampling  procedure.  In  a  fourth  stage  animal,  this 
muscle  is  1.5  mm  in  length.  Thus  each  sampling  area  is  quite  small,  and  there 
undoubtedly  was  some  heterogeneity  of  the  fiber  population  sampled  for  a  given 
area.  For  this  reason,  the  statistical  test  (Kolmogorov-Smirnoff  two-sample  test) 
was  employed  as  a  guide  rather  than  the  sole  criterion  for  determining  differences 
between  sampled  muscles. 

RESULTS 

Herrick  (1896)  reported  that  the  paired  claws  are  symmetrical  in  external 
morphology  in  the  first  three  juvenile  (postlarval)  stages  {i.e.,  stages  4  to  6)  and 
subsequently  differentiate  into  crusher  and  cutter  claws  from  stage  7  or  8  onward. 
In  stage  6  lobsters,  one  dactyl  is  always  slightly  longer  than  the  other  and  is  thus 
destined  to  be  the  cutter  claw  dactyl.  By  careful  measurement,  claw  type  in  stage 
6  could  be  unequivocally  determined.  Therefore,  the  first  three  juvenile  forms,  i.e., 
stages  4,  5,  and  6  and  several  later  stages,  namely,  stages  11,  13,  and  15,  were 
examined.  The  results  are  summarized  in  Table  I,  in  which  the  paired  closer 

TABLE  I 
Distribution  of  muscle  fiber  types  in  the  paired  claw  closer  muscles  of  juvenile  lobsters. 


Muscle  fiber  type  based  on  sarcomere  length  (^m) 

Length  of 

Stage 

animal 

Claw  I/Cutter 

Claw  II/Crusher 

(rostrum  to 

telson,  cm) 

Short 

Intermediate 

Long 

Short 

Intermediate 

Long 

4 

4-6 

6 

4 

4-6 

6 

4  (early)* 

1.2 

35% 

13', 

52  % 

32% 

15% 

53% 

4  (early)* 

1.2 

27 

8 

65 

27 

3 

70 

4* 

1.25 

35 

3 

62 

17 

3 

80 

4* 

1.25 

45 

8 

47 

32 

0 

68 

4 

1.2 

44 

0 

56 

22 

0 

78 

4 

1.2 

43 

2 

55 

32 

0 

68 

4 

1.3 

24 

0 

76 

19 

2 

79 

5  (early) 

1.4 

47 

1 

52 

27 

1 

72 

5 

1.4 

37 

0 

63 

30 

1 

69 

5  (late) 

1.5 

67 

2 

31 

24 

0 

76** 

6  (early) 

1.5 

59 

0 

41 

18 

0 

82** 

6 

1.65 

64 

1 

35 

32 

0 

68** 

6 

1.7 

58 

0 

42 

6 

? 

92** 

6 

1.7 

41 

0 

59 

27 

0 

73 

11 

3.2 

67 

1 

32 

11 

0 

89** 

11 

3.3 

79 

0 

21 

34 

2 

64** 

13 

3.9 

64 

0 

36 

0 

0 

KM)** 

15 

5.5 

82 

0 

18 

4 

0 

96** 

*  Sixty  muscle  fibers  sampled  from  each  closer  miiM-Ie;  in  all  other  animals,  90  fibers  were 
sampled  in  each  muscle. 

*  Closer  muscles  significantly  different  at  0.01  level   (Kolmogorov-Smirnov  two-sample  test). 


58 


C.  K.  GOVIND  AND  F.  LANG 


24 
20' 
16- 
12' 


CO 

en 

LU 
CD 


O 
ce 

UJ 

CD 


4- 


CLAW  I 


20-. 


16- 


12- 


4- 


CLAW  II 


q- 


34567 
SARCOMERE  LENGTH  (pm) 


10 


FIGURE  1.  Frequency  histogram  of  muscle  fibers  with  characteristic  sarcomere  lengths  from 
the  inner  aspect  of  the  paired  closer  muscles  of  a  stage  4  lobster. 

muscles  are  characterized  according  to  the  relative  distribution  of  short,  intermediate 
and  long  sarcomere  muscle  fibers.  As  in  a  previous  paper  (Lang,  Govind,  and 
She,  1977),  muscle  fiber  types  were  characterized  on  the  basis  of  sarcomere  lengths 
since  we  have  little  information  regarding  their  physiological  properties.  However, 
other  things  being  equal,  the  fibers  with  short  sarcomeres  would  contract  more 
quickly  than  fibers  with  long  sarcomeres,  just  on  the  basis  of  having  more  sarco- 
meres in  series  per  unit  length  of  fiber  (  (ahromi  and  Atwood,  1969;  Tosephson, 
1975). 

In  stages  4  and  5,  as  the  paired  claws  cannot  lie  separated  into  cutter  and 
crusher  types  from  external  morphology,  their  closer  muscles  are  labelled  as  Claw 
I  and  Claw  II  (Table  I).  In  these  cases  the  muscle  with  the  higher  percentage  of 
short  sarcomere  fibers  was  regarded  as  belonging  to  Claw  I.  In  stage  6,  and  sub- 
sequent stages,  the  paired  claws  are  externally  identifiable  by  their  dimorphic  ap- 
pearance and  their  muscles  were  classified  as  cutter  and  crusher  types  (Table  I). 


CLOSER  MUSCLE  IN  JUVENILE  LOBSTERS 


59 


Hence,  for  the  paired  muscles  in  Table  I  the  dual  heading  Claw  I/Cutter  and  Claw 
I  I/Crusher  is  used. 

In  most  of  the  early  juvenile  stages,  the  closer  muscle  was  examined  at  some 
undetermined  point  during  the  intennolt  of  that  stage.  In  several  animals  the 
muscle  was  fixed  several  hours  after  it  had  molted  into  that  stage  (early)  or  a  few- 
hours  before  it  might  have  molted  into  the  next  stage  (late). 

Stayc  -f 

At  the  molt  to  stage  4  the  lobster  assumes  its  general  adult  form  but  the  claws 
are  both  cutter-like  in  external  morphology  (Herrick,  1896,  1911).  Except 
for  animals  newly  molted  to  the  fourth  stage,  each  of  the  paired  closer  muscles  of 
fourth  stage  lobsters  is  composed  largely  of  two  distinct  populations  of  muscle 
fibers,  namely  short  sarcomere  (<  4  /Aim  and  long  sarcomere  (>  6  /Am)  fibers 
(Table  I).  The  binmdal  distribution  is  clearly  seen  in  a  frequency  histogram  of 
fiber  types  in  a  stage  4  lobster  (Fig.  1).  The  short  sarcomere  fillers  exhibit  a 
mode  at  2.5  /mi  and  long  sarcomere  fibers  at  7  /tin  ;  there  is  a  distinct  lack  of  inter- 
mediate fibers.  Intermediate  fibers  are  present,  however,  in  the  early  fourth  stage, 
where  they  make  up  approximately  10%  of  the  population.  Even  this  is  a  sig- 
nificant change  from  the  third  larval  stage  where  they  make  up  half  the  total  fiber 
population  (Lang,  Govind,  and  She,  1977).  Their  disappearance  in  the  early 
fourth  stage,  and  its  correlation  with  the  appearance  of  long  sarcomere  fibers,  will 
be  discussed  below. 

It  is  evident  that  all  stage  4  animals  examined  have  closer  muscles  with  a  sub- 
stantial number  of  short  sarcomere  fibers  (Table  I).  However,  in  no  case  did 
they  contribute  less  than  17%  or  more  than  45%  of  the  total  population.  In  this 
regard,  neither  claw  exhibits  a  closer  muscle  characteristic  of  the  adult  condition  in 
which  short  sarcomere  fibers  constitute  over  60%  of  the  population  (cutter  claw) 
or  long  sarcomere  fibers  constitute  virtually  the  entire  population  (crusher  claw). 

Owing  to  the  small  size  of  the  claws  in  this  and  other  early  postlarval  stages, 
it  is  somewhat  difficult  to  rely  on  the  data  in  regard  to  a  possible  regional  distribu- 


TABLE  II 

Regional  distribution   of  fiber  types   in   claw  closer  muscles  of  juvenile  lobsters. 


Stage 

X  umber 

Muscle  fiber  type  based  on  sarcomere  length  Gim) 

Dorsal  area 

Ventral  area 

Cutter  /Claw  I* 

Crusher  /Claw  II 

Cutter  /Claw  I 

Crusher/Claw  II 

4 

4-6 

6 

4 

4-6 

6 

4 

46 

6 

4 

4-6 

6 

4** 
4 
5 
5  (late) 
6 
11-13 

4 
3 
2 
1 
4 
4 

48% 
37 
43 
97 
84 
99 

5% 
2 

47% 
60 
55 
3 
16 
1 

40% 
49 
58 
37 

22 

7 

5% 

45% 
51 
42 
63 
78 
93 

15% 
2 
7 
27 
9 
23 

2% 

7 
1 

1 

83% 
98 
93 
66 
90 
76 

12% 
1 
3 
3 

1 

12% 

1 

76% 
99 

97 
97 
100 
98 

*  For  stages  4-6,  Claw  I  is  that  which  has  the  larger  percentage  of  fast  fibers. 
**  Claws  sampled  using  six  regions;  all  others  sampled  with  nine  regions. 


60 


C.  K.  GOVIND  AND  F.  LANG 


CO 

cc 

LU 
00 


6- 
4- 

2- 


PROXIMAL  DORSAL 


CENTRAL  DORSAL 


DISTAL  DORSAL 


_        8-1 
u_ 

PROXIMAL   MEDIAL 

U_        6- 

0 

cr      4- 

LU 
QQ        2- 

Z5 

i 

CENTRAL  MEDIAL 


DISTAL   MEDIAL 


6- 


4- 


2- 


PROXIMAL  VENTRAL 


-i     CENTRAL  VENTRAL 


10 


Uli  1  ML    VEIN 

n      [ 

1 

246        8       10 

SARCOMERE  LENGTH  (urn) 


10 


FIGURE  2.  Frequency  histogram  of  muscle  fiber  types  (based  on  sarcomere  lengths)  show- 
ing the  regional  distribution  pattern  on  the  inner  aspect  of  a  Claw  I  closer  muscle  in  a  stage 
4  lobster. 


tion  of  fiber  types.  Rather,  the  sampling  technique  is  meant  to  provide  a  survey  of 
muscle  fibers  from  all  areas  of  the  claw.  In  general,  however,  a  pattern  emerged 
that  was  consistent  among  the  seven  pairs  of  claws  examined  (Table  II).  In  the 
ventral  areas,  long  sarcomere  fibers  (sarcomeres  >  6  /mi)  comprised  88%  of  all 
fibers  sampled.  In  the  three  pairs  of  claws  where  nine  areas  were  sampled,  this 
distribution  was  even  more  striking.  Here,  where  the  three  ventral  areas  consisted 
of  the  bottom  one  third  of  the  muscle  (as  opposed  to  the  bottom  half  when  six 
areas  were  used),  long  sarcomere  fibers  comprised  over  98%  of  the  sample  (Fig. 
2).  In  contrast,  long  sarcomere  fibers  comprised  about  50%  of  the  sample  taken 
in  the  dorsal  areas  for  both  the  six  and  the  nine  region  sampling  technique. 

Stage  5 

In  stage  5,  the  claws  are  still  morphologically  identical,  but  the  latter  part  of 
this  stage  may  signal  the  transitional  period  between  the  symmetrical  claws  of 
previous  stages  and  asymmetrical  claws  of  subsequent  stages.  In  animals  from 
early  and  mid-fifth  stage,  the  muscle  fibers  again  are  largely  distributed  into  two 
distinct  populations  of  short  sarcomere  (<  4  /xm)  and  long  sarcomere  (>  6  /mi) 
fibers  (Table  I).  However,  both  claws  from  each  animal  have  fewer  than  50% 
short  sarcomere  fibers;  thus,  there  is  no  apparent  differentiation  of  claw  type.  In 
fact,  the  claws  appear  essentially  similar  to  those  in  fourth  stage  animals  with  the 


CLOSER  MUSCLE  IN  JUVENILE  LOBSTERS 


61 


exception  of  the  presence  of  a  larger  proportion  of  fibers  with  sarcomere  lengths  in 
the  range  of  8-1 1  /tin. 

In  the  one  animal  sampled  during  the  late  fifth  stage  there  was  a  striking 
change  in  the  population  of  muscle  fibers  in  one  of  the  claws  (Table  I).  Claw  I  of 
this  animal  contained  67%  short  sarcomere  fibers,  approximately  the  condition  of 
the  adult  cutter  claw.  Given  the  variability  of  the  fiber  populations  and  the  limited 
sample  from  the  late  fifth  stage,  a  definitive  conclusion  regarding  the  transition  of 
the  cutter  claw  must  await  further  sampling  during  this  period  of  growth. 

The  regional  distribution  of  fiber  types  within  stage  5  closer  muscles  was 
similar  to  that  observed  in  stage  4  animals.  Ventral  fibers  were  primarily  long 
sarcomere  (91%),  while  dorsal  fibers  were  about  equally  divided  between  short 
and  long  sarcomere  (Table  II).  Of  interest,  however,  is  the  observation  regarding 


28- 

CUTTER 

24- 

20- 

16- 

_ 

12- 

8- 

cr 

1    L    1 

1              '             Jkn 

|- 

o 

cc 
1  1  1 

~i     M                          n  rr              n 

1 

i                  i                 i                 1                 i                                                     if 

CO 

24-i 

CRUSHER 

— 

20- 

16- 

- 

12- 

- 

8- 

4- 

m             n     n               rrT 

Ml 

4567! 
SARCOMERE  LENGTH  (pm) 


10        11 


FIGURE  3.     Frequency   histogram   of   muscle   fibers   with   characteristic   sarcomere   lengths 
from  the  inner  aspect  of  cutter  and  crusher  closer  muscles  of  a  stage  6  lobster. 


62 


C.  K.  GOVIND  AND  F.  LANG 


10- 

8- 
6- 
4- 


PROXIMAL  DORSAL 


CENTRAL    DORSAL 


DISTAL  DORSAL 


GO 

£  ><h 

QQ 
u_ 

PR( 

DXIMAL  MEDIAL 

u_       / 
0 

CC        4- 
UJ 

^        2' 
-?• 

iili 

CENTRAL  MEDIAL 


DISTAL  MEDIAL 


PROXIMAL   VENTRAL 


6- 


2- 


IL 


CENTRAL  VENTRAL 


10 


DISTAL  VENTRAL 


2        4        6        8       10 

SARCOMERE  LENGTH  (urn) 


FIGURE  4.  Frequency  histogram  of  muscle  fiber  types  (based  on  sarcomere  lengths)  show- 
ing the  regional  distribution  pattern  on  the  inner  aspect  of  a  cutter  closer  muscle  in  a  stage  6 
lobster. 

the  distribution  of  fiber  types  in  the  stage  5  animal  sampled  just  prior  to  molt.  In 
the  dorsal  area  of  the  cutter  claw,  short  sarcomere  fibers  now  predominate,  as  in 
later  stages  (Table  II).  However,  in  the  other  claw,  the  presumptive  crusher, 
there  is  a  decrease  in  the  relative  number  of  short  sarcomere  fibers  in  the  dorsal 
area. 


Stage  6 

During  stage  6  one  of  the  pair  of  closer  muscles  usually  has  a  majority  of  short 
sarcomere  fibers,  while  the  other  has  a  majority  of  long  sarcomere  fibers  (Table  I ; 
Fig.  3).  In  addition,  careful  measurements  of  the  claw  at  this  stage  revealed  that 
the  dactyl  of  the  former  was  usually  longer  than  the  dactyl  of  the  latter,  an  in- 
variant characteristic  of  the  cutter  claw  in  all  later  stages.  Certainly  the  claw  with 
a  large  proportion  of  short  sarcomere  fibers  in  the  closer  muscle  resembles  the  adult 
cutter  claw  and  may  therefore  be  regarded  as  having  already  differentiated  into 
this  form. 

In  one  of  the  sixth  stage  animals  examined,  there  were  fewer  than  50%  short 
sarcomere  muscle  fibers  in  the  putative  cutter  (Table  I).  It  is  uncertain  whether 


CLOSER  MUSCLE  IN  JUVENILE  LOBSTERS 


63 


this  represents  the  variability  normally  present  in  daw  development  or  whether  it 
merely  represents  sampling  variability.  From  the  available  evidence,  the  latter 
seems  a  likely  possibility.  The  only  significant  regional  variation  for  this  claw 
occurred  in  the  medial  region  where  the  sample  revealed  equal  distribution  between 
short  and  long  sarcomere  fibers.  Among  the  other  three  sixth  stage  cutter  claws, 
the  medial  region  invariably  contained  at  least  twice  as  many  short  sarcomere  fiber  ~ 
as  long  sarcomere  fibers  (Fig.  4). 

As  in  previous  stages,  the  ventral  regions  of  both  claws  are  primarily  conipo.M-d 
of  long  sarcomere  fibers  (Table  II).  However,  in  stage  6,  there  is  a  striking- 
change  in  the  distribution  of  fibers  in  the  dorsal  regions.  In  the  cutter 
claw  the  majority  (84 ^c)  of  dorsal  fibers  have  short  sarcomeres.  while  in  the 
crusher  claw  the  majority  (78%}  of  fibers  have  long  sarcomeres.  Thus,  the  pat- 
tern of  distribution  of  fiber  types  clearly  resembles  the  adult  pattern  for  the  cutter 
claw  (Lang,  Costello  and  Govind,  1977),  while  that  for  the  crusher  claw  is  ap- 
proaching the  adult  distribution. 


28-i 


20- 

16- 
12- 


cr 

UJ 

CD 
LI_ 

u_ 
O 

tr 

UJ 

CO 


4- 


CUTTER 


n 


-P- 


20- 
16- 
12- 


4- 


CRUSHER 


4567* 
SARCOMERE  LENGTH  (urn) 


10 


11 


FIGURE  5.    Frequency  histogram  of  muscle  fibers  with  characteristic  sarcomere  lengths  from 
the  inner  aspect  of  cutter  and  crusher  closer  muscles  of  a  stage  13  lobster. 


64 


C.  K.  GOVIND  AND  F.  LANG 


C/) 

cc. 

LiJ 
CQ 


O 
cr 

UJ 

CD 


I0-i      PROXIMAL 

DORSAL 

8- 

6- 

4- 

2- 

1 

CENTRAL  DORSAL 


DISTAL  DORSAL 


_TL 


10- 
8- 

PROXIMAL  MEDIAL 

6- 

4- 

2- 

C 

1 

CENTRAL  MEDIAL 


DISTAL   MEDIAL 


a- 

6- 

PROXIMAL   VENTRAL 

4- 

2- 

rfl 

n         Hhn 

CENTRAL  VENTRAL 


p  .  .  n' 


DISTAL  VENTRAL 


8        10 


8       10 


SARCOMERE  LENGTH  (pm) 

FIGURE  6.  Frequency  histogram  of  muscle  fiber  types  (based  on  sarcomere  lengths)  show- 
ing the  regional  distribution  pattern  on  the  inner  aspect  of  a  cutter  closer  muscle  in  a  stage  13 
lobster. 

Stages  11-15 

The  characteristic  dimorphic  external  morphology  of  the  claws  is  discernible  by 
stage  8  or  9  and  is  very  distinct  by  stage  11.  At  this  point,  the  cutter  claw  closer 
muscle  has  assumed  the  adult  pattern  of  over  60%  short  sarcomere  fibers  (Table  I ; 
Fig.  5).  The  crusher  claw  on  the  other  hand  is  still  in  the  process  of  completing 
the  transformation  to  the  adult  pattern.  There  may  be  short  sarcomere  fibers 
present,  even  as  late  as  stage  16  (Goudey  and  Lang,  1974),  but  these  never 
amount  to  more  than  35%  of  the  total.  Indeed,  the  number  of  fast  fibers  is  usually 
small  and  in  some  animals  they  may  be  absent  completely  (Table  I). 

The  regional  distribution  first  manifested  in  stage  6  is  still  evident  (Table  II). 
The  dorsal  region  of  the  cutter  is  now  virtually  all  short  sarcomere  fibers  (Fig.  6), 
while  that  in  the  crusher  is  composed  of  nearly  all  long  sarcomere  fibers.  However, 
there  has  been  a  change  in  the  ventral  region  of  the  cutter  claw.  Short  sarcomere 
fibers  now  comprise  23%  of  the  population,  typical  of  that  found  in  the  adult  (Lang, 
Costello  and  Govind,  1977). 

DISCUSSION 

During  the  larval  (stages  1-3)  and  early  postlarval  (stages  4-5)  period,  the 
two  claws  of  the  lobster  are  identical  from  all  external  appearances,  and  it  is  not 


CLOSER  MUSCLE  IX  JUVENILE  LOBSTERS  65 

until  the  sixth  stage  that  their  asymmetry  is  evident.  The  claw  closer  muscles 
follow  a  similar  time  course  of  change  from  the  symmetrical  to  the  asymmetrical 
condition.  In  larval  animals,  the  closer  muscles  are  very  similar,  each  having  virtu- 
ally identical  muscle  fiber  populations.  The  same  is  true  in  the  first  two  postlarval 
stages  up  until  the  end  of  the  fifth  stage.  At  that  time,  or  during  the  sixth  stage, 
the  transformation  to  the  asymmetrical  state  occurs. 

In  light  of  the  timing  of  the  transformation  of  the  closer  muscles,  it  is  worth 
reconsidering  the  work  of  Emmel  (1908)  on  claw  "reversal"  in  the  lobster.  He 
showed,  and  we  have  confirmed  (in  preparation)  that  claw  type  is  not  established 
in  the  fourth  stage.  Normally  either  claw  has  an  equal  probability  of  being  a 
crusher  or  cutter.  However,  removal  of  one  claw  during  the  fourth  stage  will 
always  result  in  the  remaining  claw  developing  into  a  crusher.  Emmel  (1908) 
also  observed  that  this  was  true  during  the  early  fifth  stage  (within  a  day  or  two 
after  molting)  but  not  in  the  later  part  of  the  fifth  stage  or  thereafter.  In  the 
latter  cases,  removal  of  a  claw  did  not  influence  the  remaining  claw,  as  it  would 
become  a  cutter  or  a  crusher  with  equal  probability.  The  present  result  at  the 
late  fifth  stage  correlates  well  with  this  observation.  At  the  time  when  the  claws 
have  the  ability  to  "reverse",  they  are  essentially  symmetrical.  Just  after  the 
ability  to  reverse  is  lost,  the  muscles  become  asymmetrical,  one  assuming  the  char- 
acteristics of  the  adult  cutter  claw.  The  mechanisms  responsible  for  the  loss  of 
reversal  and  the  fixation  of  claw  type  are  unknown  but  perhaps  are  amenable  to 
experimental  analysis. 

It  is  of  interest  to  note  the  occurrence  of  and  changes  in  the  regional  distribution 
of  the  muscle  fiber  types.  Previous  studies  on  larval  and  adult  lobster  closer 
muscles  suggested  that  short  and  long  sarcomere  fibers  had  a  tendency  to  be  preva- 
lent in  certain  areas  (Lang,  Costello  and  Govind,  1977;  Lang.  Govind  and  She, 
1977).  In  larval  muscle  (Lang,  Govind  and  She,  1977)  as  in  the  fourth  stage,  the 
claws  are  essentially  similar  in  the  composition  and  location  of  muscle  fiber  types. 
Thus  in  the  fourth  stage,  the  dorsal  area  has  an  equal  proportion  of  short  and  long 
sarcomere  fibers  while  the  ventral  areas  averaged  over  90%  long  sarcomere  fibers. 
In  the  sixth  stage  and  perhaps  as  early  as  the  late  fifth  stage,  these  patterns  change 
dramatically.  In  the  cutter  claw  the  short  sarcomere  fibers  increase  in  prevalence 
until  they  comprise  virtually  the  entire  sample  from  the  dorsal  area  of  stage  11-13 
animals.  The  ventral  area  of  the  cutter  exhibits  little  change  during  this  growth 
period.  On  the  other  hand,  the  crusher  claw  muscle  fibers  exhibit  a  different 
pattern  in  the  dorsal  area.  Here,  the  short  sarcomere  fibers  present  in  stages  4  and 
5  are  replaced  by  long  sarcomere  fibers  over  the  next  6-8  molts.  The  ventral 
fibers,  which  are  long  sarcomere  fibers  in  stage  4,  remain  thus  in  subsequent  stages. 

What  influences  the  dimorphism  of  the  closer  muscles  such  that  short  sarcomere 
fibers  are  added  and  long  sarcomere  fibers  lost  in  the  cutter  muscle  and  vice  versa 
in  the  crusher  muscle?  The  two  excitatory  motor  axons  to  each  muscle  may  in- 
fluence the  differentiation  of  muscle  fiber  types  particularly  as  the  axons  are  them- 
selves differentiated  into  a  fast  and  a  slow  (  \Viersma,  1955,  1961)  ;  the  former  has 
a  larger  diameter  and  hence  a  faster  conduction  velocity  than  the  latter.  In  the 
cutter  claw  the  fast  axon  evokes  rapid  (20-40  msec)  closure  of  the  claw  with  a 
single  stimulus,  while  the  slow  axon  causes  a  tonic  contraction  only  at  higher 
frequencies  of  stimulation  (Wiersma,  1955;  Govind  and  Lang,  1974).  The  fast 
and  slow  axons  in  the  crusher  claw  are,  however,  "slower"  versions  of  their 


66  C.  K.  GOVIND  AND  F.  LANG 

counterparts  in  the  cutter  claw  so  that  the  fast  axmi  cannot  evoke-  a  mechanical 
response  to  single  stimuli  but  can  produce  a  small  twitch  (500  msec)  to  a  pair  of 
stimuli  (Govind  and  Lang,  1^74).  There  is,  thus,  some  correspondence  between 
the  type  of  motor  axons  and  muscle  fiber  composition  in  each  closer  muscle.  Tin- 
adult  cutter  closer  muscle  has  a  bimodal  distribution  of  fast  and  slow  muscle  fibers 
which  matches  the  fast  and  slow  axons.  The  crusher  muscle  has  a  unimodal  dis- 
tribution of  slow  muscle  fibers  which  matches  the  "slower"  versions  of  fast  and  slow 
axons  in  this  muscle.  The  differentiation  of  muscle  fiber  types  may  therefore  be 
influenced  by  its  innervating  motor  axon  through  some  type  of  neurotrophic  in- 
fluence as  has  been  demonstrated  in  vertebrate  muscle  (for  reviews  see  Guth,  1968; 
Harris,  1974;  Gutmann,  1976). 

Considering  the  striking  differences  in  morphology  of  the  chelipeds  and  the 
physiological  properties  of  their  closer  muscles,  it  is  evident  that  the  claws  must  be 
used  for  different  behaviors.  This  is  true  both  for  the  pair  of  claws  in  the  adult  as 
well  as  for  the  claws  in  larval  and  juvenile  animals  as  compared  to  the  adult  (Lang, 
Govind,  Costello  and  Greene,  1977).  Thus,  it  would  be  of  interest  to  determine 
the  physiological  properties  of  the  motor  neurons  controlling  the  chelipeds  during 
growth.  Studies  in  this  direction  are  in  progress. 


We  thank  Joseph  She  for  expert  technical  assistance,  John  Hughes  for  providing 
larval  lobsters,  and  Walter  J.  Costello  for  helpful  comments  on  the  manuscript. 
Supported  by  grants  from  N.R.C.  and  Muscular  Dystrophy  Association  of  Canada 
(to  C.K.G.)  and  NIH-NINCDS  and  Muscular  Dystrophy  Association  of  America 
(toF.L.). 

SUMMARY 

1.  The  two  chelipeds  of  the  adult  lobster  are  asymmetrical  with  respect  to  their 
external  morphology,  neuromuscular  physiology  and  utilization  in  behavior;  how- 
ever, they  are  not  genetically  fixed  in  terms  of  placement  or  handedness. 

2.  The    differentiation   of   muscle   fiber   types    was   studied   in    the   cutter   and 
crusher  claw  closer  muscles  in  the  early  juvenile  stages  of  the  lobster  Honuinis 
aincricunits.    Muscle  fibers  were  characterized  on  the  basis  of  sarcomere  length. 

3.  In  contrast  to  the  adult  lobster,  where  the  claw  closer  muscles  are  asymmetric, 
the  closer  muscles  of  the  stage  4  lobster  are  nearly  symmetric;  both  short  and  long 
sarcomere  muscle  fibers  are  present  in  each   claw  and  both   fiber  types  have  an 
identical  regional  distribution  within  the  closer  muscle. 

4.  By  stage  6  one  of  the  muscles  differentiates  into  a  cutter  muscle  with  over 
60%  short  sarcomere  fibers  and  a  distinct  regional  distribution  of  short  and  long 
sarcomere  fibers.     The  other  claw  closer  muscle  slowly  loses  its  short  sarcomere 
muscle  fibers  and  is  transformed  into  a  crusher  claw,  usually  by  stage  13-15. 

5.  The  change  of  the  closer  muscles  from  a  symmetric  to  an  asymmetric  condi- 
tion is  correlated  with  the  loss  in  ability  for  the  claws  to  undergo  a  "reversal" 
rather  than  with  the  external  appearance  of  the  claw  which  becomes  differentiated 
several  molts  later. 


CLOSER  MUSCLE  IX  JUVENILE  LOBSTERS  67 

LITERATURE  CITED 

DARBY,  H.   H..   1934.     The  mechanism   of  a-ymmetry  in  the  Alplieidae.     ('arnciiie   fust.    U'ash. 

Pul'L.  435:  347-361. 
EMMF.L.    X.    E..    1908.      The    experimental    control    of    asymmetry    at    different    stages    in    the 

development  of  the  lobster.     /.  E.rp.  Zool..  5:  471-484. 
GOUDEY,  L.  R.,  AXD  F.  LANG,  1974.     Growth  of  crustacean  muscle:  asymmetric  development  of 

the  claw  closer  muscles  in  the  lobster,  Honiarus  americanus.    J.  E.rp.  Zool.,  189:  421- 

427. 
GOVIXD,  C.  K.,  AND  F.  LAXG.  1974.     Neuromuscular  analysis  of  closing  in  the  dimorphic  claws 

of  the  lobster  Honiarus  americanus.    J.  E.rp.  Zool..   190:    281-288. 

GCTH,  L.,  1968.     "Trophic"  influences  of  nerve  on  muscle.    Physio/.  7\.v.,  48:    645-687. 
GUTMAXX,  E.,  1976.     Xeurotrophic  relations.     Annu.  AYr.  Physio!..  38:   177-216. 
HAMILTON,    P.    V.,    R.    T.    NISHIMOTO,    AND   J.    G.    HALUSKY,    1976.      Cheliped    laterality    in 

Calliiicctcs  sapid  us   (Crustacea:    Portunidae).    BioL  Bull.,  150:  393-401. 
HARRIS,   A.   J.,   1974.     Inductive   functions  of  the  nervous   system.     Annu.   Rev.   Phvsiol.,   36: 

251-305. 
HERRICK.   F.   H.,   1896.     The  American  lobster :  a  study  of  its  habits  and  development.     Bull. 

U.  S.  Fish.  Coimn.,  15:  1-252. 

HERRICK,  F.  H.,  1911.  Natural  history  of  the  American  lobster.    U.  S.  Bur.  Fish..  29:   149-408. 
HUGHES,  J.  T.,  R.  A.  SCHLESER,  AND  G.  TCHOBANOGLOUS,   1974.  A  rearing  tank  for   lobster 

larvae  and  other  aquatic  species.    Prog.  Fish.  Cult..  36:  129-132. 
JAHROMI,  S.  S.  A,XD  H.  L.  ATWOOD,  1969.     Correlation  of  structure,  speed  of  contraction,  and 

total  tension  in  fast  and  slow  abdominal  muscle  fibers  of  the  lobster,   (Honiarus  ameri- 
canus}.   J.  E.rp.  Zool.,  171 :  25-38. 
JAHROMI,  S.  S.,  AXD  H.  L.  ATWOOD,  1971.     Structural  and  contractile  properties  of  lobster  leg 

muscle  fibers.     /.  E.rp.  Zool.,  176:  475-486. 
JOSEPHSON,    R.    K.,    1975.      Extensive    and    intensive   factors    determining    the    performance    of 

striated  leg  muscle.    /.  E.rp.  Zool..  194:  135-154. 

LANG,  F.,  1975.     A  simple  culture  system  for  juvenile  lobsters.     Aqnacultitrc.  6:  389-393. 
LAXG,   F.,   W.   J.    COSTELLO,   AXD   C.   K.   GOVIXD,    1977.      Development   of   the   dimorphic   claw 

closer  muscles  of  the  lobster  Honiarus  americanus:    I.  Regional  distribution  of  muscle 

fiber  types  in  adults.    BioL  Bull..  152:  75-83. 
LAXG,  F.,  C.  K.  GOVIXD,  AXD  J.  SHE,  1977.     Development  of  the  dimorphic  claw  closer  muscles 

of  the  lobster  Honiarus  americanus:    II.   Distribution  of  muscle  fiber  types   in  larval 

forms.    BioL  Bull.,  152:  382-391. 
LAXG,  F.,  C.  K.  GOVIXD,  \Y.  J.  COSTELLO,  AXD  S.  I.  GREENE,  1977.     Developmental  neuroethol- 

ogy :  changes  in  escape  and  defensive  behavior  during  growth  of  the  lobster.     Science, 

197 :  682-685. 
LEWIS,  J.  E.,  1969.    Reversal  of  asymmetry  in  L'alappa  Weber,  1795   (Decapoda:    Oxystomata). 

Proc.  BioL  Soe.  H'ashinnton,  82:  63-80. 
PRZIBRAM,  H.,  1931.     Connectin//  la:es  in  animal  morphology.   University  of  London  Press,  Ltd., 

London,  62  pp. 
RITZMAN,  R.  E.,  1974.     Mechanisms  for  the  ^napping  behavior  of  two  alpheid  shrimp.  Alpheus 

californiensis  and  Alphcus  heterochelis.    J.  Com  p.  Physio!.,  95:  217-236. 
SHOUP.  J.  B.,  1968.     Shell  opening  by  crabs  of  the  genus  Calappa.    Science.  160  :  887-888. 
\YARXER,  G.  F.,  AXD  A.  R.  JOXES.  1976.     Leverage  and  muscle  type  in  crab  chelae   (Crustacea: 

Brachyura).    /.  Zool.  Lund..  180  :  57-68. 
WIERSMA,  C.  A.  G.,  1955.    An  analysis  of  the  functional  differences  between  the  contractions  of 

the  adductor  muscles  in  the  thoracic  legs  of  the  lobster  Honiarus  americanus.    Arch. 

Xccrland.  Zool.,  11  :  1-13. 
\YIERSMA.   C.  A.  G.,   1961.     The  neuromuscular  system.     Pages   191-240  in   T.   H.  \Yaterman, 

Ed.,  The  physiology  of  Crustacea,  I  'ol.  II.    Academic  Press,  New  York. 


Reference:  />'/<>/.  />/<//..  154:  (.8-78.  (February,  1978) 


THE  ANNUAL  REPRODUCTIVE  CYCLE  OF  AN  APODOUS  HOLO- 

THURIAN.  LEPTOSYNAPTA  TEN U IS:  A  P,l  M(  )DAL 

I',  REEDING  SEASON  x 

JEFFREY  1).  GRKKX  - 

Department  of  Zoology,  University  of  North  Carolina,  Chapel  Hill,  North  Carolina  27514 

Although  representatives  of  the  class  Holothuroidea  display  interesting  and  di- 
verse reproductive  habits  (Hyman,  1955),  their  seasonal  reproductive  biology  has 
received  little  direct  investigation.  Studies  have  been  done  on  the  aspidochirotes, 
Stichopus  japonicus  (Tanaka,  1958),  Holothnria  florldana  and  //.  inc.ricana  (Eng- 
strom,  1970),  and  H.  scabra  (Krishnaswamy  and  Krishnan,  1967)  ;  the  dendro- 
chirotes,  Sclerodactyla  (  =  Thyone  )  briarens  (Turner,  1966)  and  Cncinnaria  psen- 
docurata  (Rutherford,  1973);  and  the  apodid.  Rhabdomolgns  rubcr  ( Menker, 
1970).  However,  only  the  studies  by  Tanaka  (1958)  and  Engstrom  (1970)  have 
presented  detailed  histological  data  concerning  the  annual  gonadal  cycle.  There 
appears  to  be  no  similar  histological  study  of  reproduction  in  any  apodous  holo- 
thurian. 

McCrary  (1969)  studied  the  seasonal  occurrence  over  a  three-year  period  of  the 
planktonic  larvae  of  Leptosynapta  tennis  (Ayres),  a  common  synaptid  of  the  east 
coast  of  the  United  States.  She  concluded  that  this  species  had  "two  discrete 
breeding  seasons"  on  the  North  Carolina  coast.  The  present  study  describes  the 
bimodal  reproductive  cycle  and  gonad  histology  of  a  North  Carolina  population  of 
Leptosynapta  tennis  and  the  relationship  between  spermatogenesis  and  oogenesis  in 
this  hermaphroditic  (  Costello  and  Henley.  lc>71  )  holothurian. 

MATERIALS  AND  METHODS 

Specimens  of  Leptosynapta  tennis,  an  infaunal  species,  were  collected  by  digging 
in  the  sediments  at  Wrightsville  Beach,  North  Carolina,  at  approximate  monthly 
intervals  from  June,  1972,  through  August,  1973,  with  an  additional  collection  in 
March,  1974.  At  each  collection  4  to  25  holothurians  were  taken.  The  sample 
sizes  were  influenced  by  the  quality  of  the  low  tides,  and  therefore  numbers  of  col- 
lected individuals  varied  from  month  to  month.  However,  only  4  of  the  17  samples 
produced  less  than  10  specimens. 

Whole  gonads  were  fixed  in  Benin's  solution  (Galigher  and  Kozloff,  1971)  or 
in  Susa's  fixative  (Humason,  1962).  Paraffin  (56-58°  C,  melting  point)  or  Steed- 
man's  polyester  (Steedman,  I960)  were  used  for  embedding.  Sections  were  cut  at 
6  to  10  ,um  and  stained  in  Heidenhain's  hematoxylin  (Galigher  and  Kozloff,  1971). 

Twenty  to  fifty  post-pachytene  oocytes  encountered  sequentially  in  each  sec- 
tioned gonad  were  measured  in  nucleolar  section  in  order  to  determine  their  size 

1  This  work  is  part  of  a  thesis  submitted  in  partial  fulfillment  of  the  requirements  for  the 
degree  of  Master  of  Arts. 

~  Present  address :  Department  of  Anatomical  Sciences,  State  University  of  New  York  at 
Buffalo,  Buffalo,  New  York  14214. 

68 


REPRODUCTION  IN  LEPTOSYNAPTA   TEXUIS  69 

frequencies.  The  position  of  each  measured  oocyte  was  carefully  noted  to  avoid 
measuring  the  same  cell  twice.  Since  most  of  the  cells  were  somewhat  elliptical 
in  outline,  the  diameter  was  calculated  by  averaging  the  long  and  short  axes 
(Holland,  Grimmer,  and  Kubota,  1975).  Xonsectioned  gonads  were  observed 
through  a  dissecting  microscope,  oocytes  being  measured  with  an  ocular  microme- 
ter. Mean  oocyte  diameters  were  calculated  for  each  animal,  ranked  in  groups  at 
10-micron  intervals  from  less  than  20  /xm  to  200  /xm  (at  which  size  they  are 
spawned),  and  used  to  construct  average  frequency  polygons  for  each  collection  date. 
The  polygons  were  shaded  to  show  the  relative  contribution  of  sectioned  oocyte 
measurements  to  the  total  sample.  The  data  were  pooled  to  determine  the  mean 
oocyte  diameter  and  standard  deviation  for  each  collection. 

The  testicular  portion  of  each  gonad  was  assigned  to  one  of  five  stages  based  on 
cell  types  present,  and  a  gonad  maturity  index  (MI),  similar  to  one  used  by 
Yoshida  (1952)  for  work  on  the  echinoid  Diadeuia  setosiini,  was  devised  here  for 
testicular  maturity  of  the  ovotestis  of  L.  tennis.  The  index  is  calculated  by  the 
formula,  MI  =  [1  (number  of  animals  in  stage  I)  +  2  (number  of  animals  in 
stage  II)  +  .  .  .]/  total  number  of  animals  staged  that  month. 

Mean  weekly  sea  water  temperatures  were  furnished  through  the  courtesy  of  the 
International  Nickel  Company,  Inc.,  at  \Yrightsville  Beach. 

RESULTS 
Gonad  morphology 

The  gonad  of  Leptosynapta  tennis  is  composed  of  dichotomously  dividing  tubu- 
les. The  degree  of  branching  depends  on  the  reproductive  condition  of  the  animal. 
In  reproductively  active  animals  the  tubules  are  proliferated,  while  in  inactive  or 
spawned  out  individuals  the  tubules  are  shrunken  and  only  a  few  branches  re- 
main. The  right  and  left  halves  of  the  gonad  unite  in  the  dorsal  mesentery,  and 
the  ciliated  gonoduct  is  short,  opening  to  the  outside  through  a  small  papilla  situ- 
ated between  two  dorsal  tentacles. 

In  L.  tennis  there  appears  to  be  no  specialization  in  gonadal  tubules  to  contain 
only  male  or  female  sex  cells.  Testicular  tissue  is  present  between  maturing  oocytes 
and  most  tubules  contain  abundant  testicular  tissue  as  well  as  ovarian  tissue. 

O  agenesis 

The  gonads  of  72  animals  were  sectioned ;  all  contained  post-pachytene.  primary 
oocytes  at  least  10  /xm  in  diameter,  with  a  distinct  nucleolus  in  a  large  germinal 
vesicle  (Figs.  1-6).  During  the  course  of  oogenesis  several  nuclear  and  cytoplasmic 
changes  occurred.  Presumed  oogonia  were  characterized  by  their  small  diameter 
(less  than  10  /xm)  and  their  strongly  basophilic  nuclear  granules  (Fig.  2).  These 
cells  lacked  a  single,  distinct  nucleolus.  Presumably,  after  a  number  of  mitotic 
divisions  (Fig.  3),  oogonia  gave  rise  to  primary  oocytes.  The  primary  oocytes 
underwent  a  series  of  chromatin  transformations  during  the  meiotic  prophase  stages 
of  leptonema  through  pachynema  (collectively  called  the  spireme  stages).  In 
these  stages  the  chromosomes  were  distinct  because  they  had  condensed  and  were 
seen  as  thick,  individual  threads  in  the  clear  nucleoplasm  (Fig.  1).  In  the  later 


70 


JEFFREY  D.  GREEN 


%• .  / 

m  ^*     ')( 


i* 


r-  ^J^  •  VA  W^W  *^Bi  "I*1    "^  "*¥» 


FIGURES  1-8.  Heidenliain's  hematoxylin  stained,  sectioned  gonads  of  Leptosynapta  tennis 
showing  stages  in  oogenesis. 

FIGURE  1.  Spireme  oocyte  (arrow)  and  young  post-pachytene  oocytc  (twin  arrows)  with 
prominent  nucleolus.  Scale  is  10  /un. 

FICURE  2.  Oogonia  (arrow),  oocytes  (ooc),  and  follicular  cell  nuclei  (twin  arrows); 
stage  II  testicular  tissue.  Scale  is  20  /j.m. 

FIGURE  3.     Mitotic  oogonium  (arrow).     Scale  is  10  MI"- 

FIGURE  4.     Dij)lotene  oocytes;  stage  III  testicular  tissue.     Scale  is  50  /j.m. 

FIGURE  5.  Small  post-pachytene  oocyte  (arrow)  in  a  May  31  specimen.  The  large  oocyte 
has  diffuse,  homogeneous  chromatin  in  the  germinal  vesicle ;  stage  I  testicular  tissue.  Scale  is 
50  ,um. 

FIGURE  6.    Nucleolus  in  germinal  vesticle  ;  phase  contrast.    Scale  is  10  /um. 

FIGURE  7.  Follicular  membrane  (arrow)  surrounding  oocytes;  phase  contrast.  Scale  is 
50  /tm. 

FIGURE  8  .  Ruptured  follicular  membrane  (arrow)  indicating  recent  shedding  of  mature 
oocytes;  stage  II  testicular  tissue.  Scale  is  50  /j.m. 


REPRODUCTION   IX  LEPTOSYX Al'T.l    TENUIS  71 

spireme  stages  a  "bouquet"  configuration,  in  which  the  chromosomes  were  arranged 
in  the  nucleus  with  their  free  ends  directed  toward  one  pole,  was  sometimes  ob- 
served. Subsequently,  the  thick  chromatin  threads  decondensed,  the  nucleolus 
grew  prominent,  the  germinal  vesicle  enlarged,  and  the  cell  became  a  post-pachytene 
oocyte.  These  young  oocytes  had  a  large  germinal  vesicle  approximately  65%  of  the 
diameter  of  the  cell,  and  a  nucleolus  approximately  20%  of  the  cell's  diameter.  The 
chromatin  was  strongly  basophilic  and  apparently  double  stranded  (diplonema) 
(Fig.  4).  As  the  oocytes  grew,  both  the  nucleus  and  nucleolus  enlarged  more 
slowly,  so  that  the  final  ratios  of  nucleus  and  nucleolus  diameters  to  oocyte 
diameter  were  approximately  50%  and  10%,  respectively.  Also,  large  oocytes 
were  characterized  by  a  diffuse,  granular,  and  lightly  staining  chromatin  (Fig.  5). 
Large  nucleoli  clearly  showed  an  outer  basophilic.  vesicular,  cup-shaped  portion 
and  a  lighter  inner  portion  (Fig.  6).  The  nucleolus  was  eccentric  and  rested 
against  the  nuclear  membrane. 

Xo  further  changes  were  observed  in  the  morphology  of  the  oocytes,  which  were 
spawned  at  approximately  200  /mi.  Maturation  divisions  take  place  during  spawn- 
ing and  fertilization  in  holothurians  (Kume  and  Dan,  1968),  therefore  stages  later 
than  primary  oocytes  were  not  present  in  the  gonads  of  L.  tennis. 

Developing  oocytes  were  arranged  in  single  layers  and  surrounded  by  a  flat 
follicular  epithelium  (Fig.  2)  with  prominent  ovoid  nuclei  3  by  6  /*m  in  diameter. 
As  the  oocytes  grew,  they  bulged  into  the  gonadal  lumen,  carrying  with  them  their 
follicular  epithelium  (Fig.  7).  Eventually  oocytes  broke  free  from  their  surround- 
ing epithelium  (Fig.  8),  presumably  through  the  muscular  activity  of  the  gonadal 
wall  and  were  shed. 

Spermatogenesis 

Testicular  tissue  was  distinguishable  in  48  of  the  72  animals  sectioned,  although 
developing  oocytes  were  present  in  all  of  them.  Those  which  did  not  exhibit 
spermatogenic  stages  were  either  spawned  out  individuals  or  animals  collected  dur- 
ing the  winter  months  when  gametogenesis  was  at  a  low  point.  Presumably, 
spermatogonia  were  present  throughout  the  year,  but  were  possible  to  distinguish 
from  oogonia  only  by  their  position  within  testicular  tissue  and  only  during  the 
testicular  growth  phase  starting  in  early  spring.  Spermatogonia  (Fig.  9)  were  10 
/Ain  in  diameter  with  a  nucleus  approximately  6  to  7  ^m  in  diameter,  having  granular 
chromatin.  These  spermatogonia  gave  rise  to  primary  spermatocytes  (Fig.  9), 
which  contained  a  nucleus  5  /xm  in  diameter  with  darkly  staining  granular  chroma- 
tin.  Occasionally,  primary  spermatocytes  were  seen  in  metaphase  preparing  to 
divide  into  secondary  spermatocytes.  However,  most  of  the  spermatocytes  observed 
were  similar  in  size  and  appearance ;  these  were  probably  primary  spermatocytes. 
Secondary  spermatocytes  and  spermatids  were  not  observed  in  the  same  numbers 
as  primary  spermatocytes,  presumably  because  those  stages  were  of  short  duration. 
The  spermatozoon  (Fig.  9)  had  a  rounded  head  approximately  2  to  3  ju,m  in  diame- 
ter, a  small  midpiece,  and  a  flagellum  which,  in  gonad  smears,  appeared  to  be  50 
/mi  long. 

To  quantify  testicular  development,  the  testicular  portion  of  each  gonad  was 
assigned  to  one  of  five  stages.  In  stage  I,  sperm  had  been  shed,  although  a  few 


72 


JEFFREY  1).   CiREEN 


;. ..  ;s.  'V  \  :  ;•« 


FIGURES  9-14.  Heidenhain's  hematoxylin  stained,  sectioned  gonads  of  Leptosynapta  tennis 
showing  testicular  stages  and  spermatogenesis. 

FIGURE  9.  Testicular  region  of  gonad  with  presumed  spermatogonia  (arrow),  spermato- 
cytes  (spc),  and  spermatozoa  (twin  arrows).  Scale  is  10  /mi. 

FIGURE  10.  Stage  I  testicular  region  with  relict  spermatozoa  in  the  gonadal  lumen.  Scale 
is  20  /urn. 

FIGURE  11.    Gonad  in  stage  II  testicular  condition;  gonadal  lumen  (him).    Scale  is  50  /xm. 

FIGURE  12.  Stage  III  testicular  tissue  (arrow)  between  oocytes.  First  indication  of  sperm- 
atogenesis for  1973  occurred  in  February.  Scale  is  50  Mm. 

FIGURE  13.  Stage  IV  testicular  tissue  (arrow).  Spermatozoa  have  not  yet  appeared  in 
the  gonadal  lumen.  Scale  is  50  ,um. 

FIGURE  14.     Stage  V  testicular  tissue   (arrow).     Spermatozoa  fill  the  lumen.     Scale  is  50 


residual  sperm  may  have  remained  in  the  lumen  (Fig.  10),  and  the  spermatocyte 
layer  had  disappeared.  In  stage  II,  the  testicular  components  of  the  gonad  were 
inconspicuous  (Fig.  11).  Any  spermatogonia  which  may  have  been  present  were 
not  distinguishable  from  oogonia.  Beginning  spermatogenesis  marked  stage  III 
(Fig.  12),  in  which  there  were  clumps  of  spermatogonia  and  spermatocytes  two  or 
three  cells  thick  scattered  around  the  gonad  periphery  between  oocytes.  Advanced 
spermatogenesis,  with  a  thick  spermatocyte  layer  of  intensely  stained  cells,  was 
characteristic  of  stage  IV  (Fig.  13).  Spermiogenesis,  however,  was  not  very 
advanced,  as  evidenced  by  a  patent  lumen.  Stage  V  was  the  mature  stage  (Fig. 
14).  A  thick  spermatocyte  layer  persisted,  but  the  lumen  was  packed  with  sperm- 
atids  and  spermatozoa.  Gonad  smears  produced  actively  swimming  sperm.  As 
spermatogenesis  proceeded  in  stage  V,  the  spermatocyte  layer  was  reduced,  so  that 


REPRODUCTION  IX  LEPTOSYNAPTA   TENUIS 


73 


following  spawning  only  small  areas  remained  identifiable  as  testicular  tissue,  and 
stage  I  was  restored. 

Comparisons  within  each  individual  between  testicular  development  and  mean 
oocyte  diameter,  regardless  of  the  time  of  year,  showed  a  clear  relationship  (Fig. 
15).  Animals  with  stage  II  testicular  tissue  had  small  oocytes  averaging  30  /mi  in 
diameter.  Stages  III  and  IV  were  found  in  animals  with  an  average  oocyte  diame- 
ter of  38  and  51  /mi,  respectively.  A  wide  range  of  oocyte  diameters  (60  to  160 
/mi)  was  present  along  with  stage  Y  testicular  tissue.  Most  animals  with  stage  I 
testicular  regions  had  a  mean  oocyte  diameter  of  23  /mi,  although  two  animals  also 
had  oocytes  larger  than  120  /mi  in  diameter.  These  latter  probably  had  spawned 
their  male  gametes,  but  still  had  not  spawned  all  their  oocytes. 

These  last  data  suggest  that  Leptosynapta  tennis  is  a  simultaneous  hermaphro- 
dite, developing  and  shedding  both  sperm  and  oocytes  within  the  same  reproductive 
period.  Oocytes  were  present  throughout  the  year,  but  spermatogenic  stages  seemed 
to  be  transitory.  It  appears  that  each  animal  underwent  spermatogenesis  rather 
quickly,  so  that  sperm  were  shed  when  the  oocytes  were  large  but  still  growing 
(Fig.  15).  The  oocytes  were  then  shed  somewhat  later. 

Annual  reproductive  cycle 

In  1972-1973,  Leptosynapta  tennis  bred  at  two  different  times  of  the  year  at 
\Yrightsville  Beach,  North  Carolina.  This  population  spawned  in  the  spring  and 
again  in  the  fall,  as  indicated  by  the  bimodal  occurrence  of  mature  primary  oocytes 
over  the  year  (Fig.  16).  Mean  oocyte  diameters  of  the  population  reached  peaks 
in  June  and  October,  1972,  and  May  and  August,  1973.  The  development  of 


o 
Q 


O 

o 
O 

c. 
o 

0> 


200  r 


160 


120 


80 


40 


n  nr  n  i 

Testis  Stage 

FIGURE  15.  Comparison  between  ovarian  and  testicular  development  in  the  ovotestis  of  each 
animal:  y=  -9.69  +  18.38x,  n  -  61 ;  r  -  0.79,  P<0.01.  (The  two  individuals  in  testis  stage 
I  having  oocytes  of  128  /tm  and  190  /j.m  were  omitted  from  the  correlation  and  regression  calcu- 
lations.) These  data  are  from  sectioned  material  only. 


74 


JEFFREY  D.  GREEN 


°c 

32-1 
28- 
24- 

20- 

16- 

12- 

8- 

4- 


51 
4- 

3- 
2- 
1- 


Mean  Sea  Temperature 


Testicular  Maturity 
Index 


urn 

200-, 

,  1QQ%.                          Oocyte    Diameter                                    10                ,, 

1.5                                                                                       25                        n                          £ 

\           4                                       J 

13 

11 

150- 

I 

-7 

| 

" 

1     - 

/ 

, 

\ 

25          7    / 

~~_. 

15 

100- 

X 
X 

/ 

x'>». 

\     J 

fl/ 

X 
X 

\ 

x 

\                                                / 

\                              18 

/ 

\  . 

X 

\ 

N                            19               I 

1       6     . 

\-- 

1 

, 

\ 

50- 

- 

-i     sx               / 

) 

\^io    -R__: 

^TS 

) 

I 

\12,' 

\ 

\ 

J           ^^^ 

^r"-<:  - 

n\ 

/     - 

1 

'^- 

_A 

•^ 

\y 

9           9             23                   21       20        22       19         16     6    17        14    3        31        30       29       27          '8 

J           J                ASONDJFM              AM                J           J            A        Mar 

1972                                                                                       1973                                                                                                1974 

FIGURE  16.  Average  size-frequency  polygons  of  post-pachytene  oocytes  in  the  population. 
Each  polygon  was  constructed  using  the  mean  oocyte  diameters  from  the  number  of  individuals 
shown  above  the  polygon.  Combined  measurements  of  sectioned  and  whole  oocytes  were  used 
to  construct  the  polygons  which  are  shaded  to  show  the  relative  contribution  of  sectioned  oocyte 
measurements  to  the  total  sample.  Population  mean  oocyte  diameters  are  shown  by  the  broken 
line.  Standard  deviations  for  most  collections  are  indicated  by  horizontal  lines  intersecting  the 
polygons.  Testicular  maturity  indices  (with  standard  deviations  shown)  are  correlated  with 
the  mean  oocyte  diameters  for  the  population  (r  —  0.84,  P<0.01).  Mean  weekly  sea  tempera- 
tures are  also  plotted. 

testicular  regions  of  the  gonad  also  followed  a  bimodal  pattern  throughout  the  year 
(Fig.  16)  and  correlated  well  with  the  mean  oocyte  diameters  (r  —  0.84,  P  <  0.01). 
A  mid-summer  lowr  of  reproductive  activity  is  indicated  by  the  absence  of  mature 
oocytes  from  the  August,  1972,  sample  and  the  absence  of  both  mature  oocytes  (Fig. 
16)  and  spermatozoa  (Table  I  )  from  the  July,  1973,  collection.  (Mature  gametes 
were  also  absent  from  December,  1972,  through  March,  1973. )  These  findings 
support  the  conclusion  of  McCrary  (1969),  based  on  larval  occurrence  in  the 
plankton,  that  this  species  has  two  discrete  breeding  periods  each  year. 

A  striking  difference  existed  between  the  range  of  oocyte  diameters  in  the 
population  for  most  collections  and  the  range  present  in  single  individuals.  Al- 
though a  full  range  of  oocyte  diameters  occurred  in  the  population  during  the 
spawning  periods,  no  individuals  had  such  a  range.  On  the  average,  each  animal's 
oocyte  diameters  were  within  a  range  of  ±23%  of  its  mean  oocyte  diameter  (Green, 
1976).  A  few  animals  (two  in  November,  two  in  May,  and  one  in  August,  1973) 
had  two  size  classes  of  oocvtes  within  the  same  gonadal  tubules.  The  smaller 


REPRODUCTION  IN  LEPTOSYNAPTA   TENUIS 


75 


oocytes  were  at  least  100  /mi  smaller  than  the  large  oocytes,  except  in  one  specimen 
from  November,  where  the  difference  was  approximately  60  /mi  between  size 
classes.  However,  the  small  oocytes  in  these  cases  represented  a  small  part  of  the 
total  oocytes  observed  and  were  found  singly  near  larger  oocytes  (Fig.  5).  The  fate 
of  the  small  oocytes  (that  is,  whether  or  not  they  reach  maturity  or  degenerate)  is 
not  known. 

One  animal  from  the  October,  1972,  collection  appeared  to  have  recently 
spawned  some  of  its  oocytes.  Some  of  its  gonadal  tubules  were  empty,  while  others 
had  fully  developed  oocytes  present  in  single  file  in  proximal  regions  of  the  tubules. 
These  latter  tubules  were  just  wide  enough  (ca.  200  /mi)  to  accommodate  the  ma- 
ture oocytes.  During  oocyte  growth  the  tubules  may  attain  a  diameter  of  700  /mi. 
Presumably,  muscles  in  the  gonadal  walls  contract  to  push  the  oocytes  through.  This 
individual  from  October  apparently  had  not  finished  spawning  for  the  year  when  it 
was  collected.  This  observation  raises  a  question  of  whether  one  animal  completes 
its  spawning  all  at  once,  or  whether  it  spawns  several  times  during  a  specific  time 
interval.  Newly  formed  post-pachytene  oocytes  (less  than  40  /mi)  were  not  ob- 
served in  October,  perhaps  indicating  that  the  oogenic  cycle  was  involved  with 
oocyte  growth  rather  than  the  production  of  new  oocytes. 

Reproductive  activity  decreased  during  November,  and  the  small  oocytes  present 
may  have  been  representatives  of  the  1973  generation.  Unidentified  cells  along 
with  degenerating  oocytes  in  the  gonad  may  be  evidence  of  phagocytosis  of  un- 
spawned  oocytes  (Tanaka,  1958;  Engstrom,  1970).  Degenerating  oocytes  and 
unidentified  cells  were  observed  in  a  few  other  months  as  well,  but  this  observation 
was  made  so  infrequently  that  its  significance  cannot  be  judged.  In  November  the 
gonads  also  contained  dividing  spermatocytes  and  mature  spermatozoa. 

By  December  22,  no  mature  oocytes  were  observed  in  the  gonads.  Residual 
sperm  (Fig.  10)  and  debris  of  degenerating  oocytes  were  present  in  the  gonadal 
lumina.  Although  spermatogenesis  had  ceased,  there  were  many  spireme  oocytes 
representing  a  new  ovarian  cycle. 

TABLE  I 

Number  of  Leptosynapta  tennis  in  each  testicular  stage  per  sample.     Sectioned  material  from  tin' 
en/lections  of  November  21,  1972,  through  August  27,  1973,  was  used  to  calculate 
the  maturity  index. 


Date 

Stage  I 

it 

III 

IV 

V 

Maturity 
index 

±s.d. 

Nov.  21,   1972 

1 

4 

4.4 

±1.3 

Dec.  22 

5 

1.0 

±0 

Ian.  19,  1973 

2 

3 

1.6 

±0.5 

Feb.  16 

4 

1 

2.2 

±0.4 

March  6 

1 

4 

2.8 

±0.4 

March  1  7 

4 

1 

3.2 

±0.4 

April  14 

1 

3 

1 

3.8 

±1.1 

May  3 

1 

4 

4.2 

±1.8 

May  31 

1 

1 

3 

4.2 

±1.3 

)nne  30 

1 

2 

2 

4.0 

±1.2 

July  29 

7 

1 

2.1             ±0.3 

August  27 

1 

1 

1 

1 

2.5            ±1.3 

76  JEFFREY  D.  GREEN 

The  January-to-early-March  phase  of  the  reproductive  cycle  was  characterized 
by  slow  hut  steady  gametogenesis.  Oocyte  growth  occurred  slowly  through  early 
March  (Fig.  16),  the  population  mean  increasing  hy  less  than  15  /xin.  Spireme 
oocytes  were  present  throughout  this  period,  which  was  dominated  by  the  emergence 
of  new  oocytes.  Residual  spermatozoa  were  still  present  in  January,  but  were  not 
seen  in  later  winter  specimens.  By  February  there  was  evidence  of  a  new  spermato- 
genic  cycle  (Fig.  12),  but  most  testicular  regions  were  still  in  the  recovery  (II) 
stage  (Table  I). 

By  the  middle  of  March  the  "quiescent"  period  of  oocyte  growth  seemed  to  have 
ended  with  the  maximum  oocyte  diameters  reaching  100  /xin.  Most  testicular 
regions  had  meiotic  spermatocytes.  In  April,  spireme  oocytes  were  not  seen  as 
frequently  as  in  early  March,  indicating  that  growth  of  oocytes  rather  than  produc- 
tion of  new  oocytes  was  predominant.  For  the  first  time  in  1(>73,  mature  actively 
swimming  spermatozoa  were  observed  in  gonad  smears. 

The  height  of  reproductive  activity  was  reached  by  May  3.  Mature  oocytes 
and  spermatozoa  were  present,  with  some  gonads  already  showing  spawned  out 
testicular  regions.  One  animal  with  small  oocytes  (30  ju,m)  from  May  31  appeared 
to  have  recently  shed  mature  oocytes,  owing  to  the  presence  of  a  ruptured  follicular 
membrane  (Fig.  8).  Neither  May  collection  revealed  emerging  post-pachytene 
oocytes  (no  spireme  stages  observed),  oogenesis  still  being  dominated  by  growth 
of  oocytes.  Two  individuals  contained  two  distinct  oocyte  size  classes  separated  by 
more  than  100  jam.  Since  the  smaller  oocytes  were  larger  than  any  that  were  ob- 
served in  July,  the  possibility  exists  that  two  generations  of  oocvtes,  widely  sepa- 
rated in  size,  matured  and  were  spawned  within  a  single  breeding  period,  in  this 
case,  the  spring  period.  Deteriorating  oocytes  were  not  observed,  but  the  possi- 
bility that  the  second  size  class  of  oocytes  degenerated  cannot  be  eliminated  at  this 
time. 

In  addition  to  nearly  mature  oocytes,  numerous  small  oocytes  were  present  in 
June  (Fig.  16),  indicating  that  production  of  post-pachytene  oocytes  had  occurred. 
Spermatozoa  and  spermatocytes  continued  to  be  present  in  late  Tune. 

The  first  half  of  the  annual  breeding  season  apparently  ended  between  June  30 
and  July  29.  The  July  specimens  resembled  those  of  the  previous  August,  all 
animals  having  small,  thin  gonads,  approximately  one  fifth  of  their  maximum 
diameter.  Oocytes  were  less  than  30  /J.MI  in  diameter,  and  the  presence  of  numerous 
spireme  oocytes  signaled  the  proliferation  of  new  oocytes  for  the  autumn  reproduc- 
tive period.  The  animal  in  July  with  the  largest  oocytes  (mean  diameter  of  26  tun) 
also  had  the  most  advanced  testicular  stage  (III),  although  testicular  portions  in 
most  animals  were  in  the  recovery  (II)  stage  (Table  I). 

Twenty-nine  days  later  on  August  27,  a  wide  range  of  gametogenic  stages  was 
again  present  in  the  population  (Table  I  and  Fig.  16).  Some  of  these  specimens 
apparently  had  spawned  already,  and  the  reproductive  activity  of  the  population 
appeared  headed  for  a  second  peak  representing  the  autumn  breeding  period. 

Evidence  that  the  year-to-year  reproductive  patterns  are  similar  is  shown  by 
the  data  from  March,  1974  (Fig.  16)  ;  these  data  are  comparable  to  those  of  the 
previous  March. 


REPRODUCTIOX  IX  LEPTOSYNAPTA   TEXL'IS  /  / 

DISCUSSION 

Leptosynapta  tennis  appears  to  breed  twice  during  the  year  in  Xorth  Carolina, 
with  a  mid-summer  cessation  of  reproduction  occurring  in  July  or  August.  Only 
Krishnaswamy  and  Krishnan  (1967)  have  suggested  a  himodal  breeding  season  for 
any  other  holothurian.  They  reported  that  Holothnria  scahra  had  breeding  peaks 
in  July  and  October  in  southern  India.  However,  individuals  with  mature  gonads 
were  found  in  collections  between  July  and  October.  In  contrast  to  H.  scabra,  only 
young  oocytes  were  found  in  the  gonads  of  L.  tennis  during  August.  1972,  and 
July,  1973.  In  addition,  McCrary  (1969)  reported  an  absence  of  the  planktonic 
larvae  of  Leptosynapta  during  the  same  months  in  her  three-year  study  of  the 
plankton  on  the  Xorth  Carolina  coast. 

Factors  governing  gametogenesis  and  spawning  have  not  been  elucidated  for 
holothurians,  although  salinity  (Krishnaswamy  and  Krishnan,  1967)  and  tempera- 
ture (Tanaka,  1958)  have  been  proposed  as  regulatory  factors.  It  is  logical  to  as- 
sume that  a  reproductive  cycle,  such  as  L.  tennis  seems  to  exhibit,  could  be  regulated 
by  external  factors.  At  \Yrightsville  Beach,  salinity  values  were  erratic  through- 
out each  year,  but  mean  weekly  sea  water  temperatures  were  similar  in  1972  and 
1973  (Fig.  16).  Furthermore,  the  apparent  mid-summer  spawning  hiatus  coincided 
with  the  highest  temperatures  in  each  year.  The  data  of  McCrary  (  1969  )  reveal  a 
similar  relationship.  Clearly,  experimental  work  on  the  effects  of  temperature  on 
gametogenesis  in  this  species  is  needed  to  clarify  what  relationship,  if  any,  exists. 

The  assumption  of  a  bi modal  breeding  season  for  L.  tennis  requires  that  gameto- 
genic  growth,  which  occurred  over  a  five-month  period  during  the  winter  and  early 
spring,  could  also  reach  completion  in  as  little  time  as  one  month  or  less  in  sum- 
mer or  fall.  The  data  (Fig.  16)  indicate  that  this  is.  indeed,  the  case.  On  July  29, 
only  small  oocytes  (30  ^m  or  less)  were  observed  in  the  population,  but  on  August 
27,  29  days  later,  mature  oocytes  (200  /im  )  were  found.  The  increase  of  170  p.m  in 
oocyte  diameter  is  equivalent  to  a  \9.\c/c  per  day  increase  in  volume  (calculated  by 
the  instantaneous  relative  growth  rate  method  of  Hrody.  1945).  There  are  no 
data  in  the  literature  concerning  these  growth  rates  in  holothurian  oocytes.  but 
there  are  reports  of  oocyte  growth  rates  for  the  echinoid,  Strontjylocentrotits  pur- 
pnratits  (Conor,  1973)  and  the  crinoid.  Commit/ins  japonica  (Holland,  Grimmer, 
and  Kubota,  1975).  However,  these  species  required  almost  a  year  for  oocyte 
maturation,  and  they  spawned  only  during  one  short  interval  each  year.  Only  the 
winter  growth  rate  of  Leptosynapta  tennis  oocytes  (approximately  4c/(  per  day  » 
was  comparable  to  the  rates  for  5".  purpnratns  and  C.  japonica. 


I  thank  my  advisor,  Dr.  C.  E.  Jenner,  and  Drs.  R.  M.  Rieger  and  E.  A.  Mc- 
Mahan  for  their  valuable  help  and  advice  during  the  course  of  this  study.  I  also 
thank  Dr.  R.  G.  Summers  for  his  critical  reading  of  the  manuscript  and  his  sugges- 
tions. 

SUMMARY 

1.  Gonads  of  Leptosynapta  tennis  were  examined  histologically,  and  gameto- 
genesis in  this  apodid  holothurian  is  described. 


JEFFREY  I).  GREEN 

2.  L.  tennis  is  a  simultaneous  hermaphrodite.    Each  animal  produces  and  sheds 
spermatozoa  and  oocytes  during  the  same  breeding  season,  although  sperm  seem  to 
be  shed  somewhat  earlier  than  oocytes. 

3.  Testicular  maturity  indices  and  oocytc   diameter  measurements  revealed  a 
bimodal  annual   reproductive   cycle   in   a    North    Carolina   population   sampled   at 
monthly  intervals  over  a  fifteen-month  period.     Breeding  occurred  for  about  three 
months  both  preceding  and  following  a  month,  or  so,  of  breeding  inactivity  in  July 
or  August. 

4.  The  mid-summer  absence  of  mature  gametes  in  the  population  coincided  with 
the  highest  sea  water  temperatures,  suggesting  a  temperature  regulated  gametogenic 
cycle. 

5.  The  high  oocyte  growth  rate  leading  to  the  second  spawning  season  in  the 
autumn  is  perhaps  five  times  that  of  the  winter  rate. 

LITERATURE  CITED 

BRODY,  S.,  1945.    Biogenergctics  and  groii'th.    Reinhold,  New  York,  1023  pp. 

COSTELLO,  D.  P.,  AND  C.  HENLEY,  1971.     Methods  jor  obtaining  and  handling  marine  eggs  and 

embryos.  2nd  ed.    Marine  Biological  Laboratory,  Woods  Hole,  Massachusetts,  247  pp. 
ENGSTROM,   N.   A.,    1970.     The   reproductive   cycles,   systematic   status   and  general   biology   of 

Holothuria    (Halodeima)    ftoridana    Pourtales    1851    and   H.    (H.)    mc.ricana    Ludvvig 

1875.    M.A.  Thesis,  University  of  Miami,  Miami,  Florida,  92  pp. 
GALIGHER,  A.  E.,  AND  E.   N.  KOZLOFF,   1971.     Essentials  of  practical  microtechnique,  2nd  cd. 

Lea  and  Febiger,  Philadelphia,  531  pp. 
CONOR,   J.  J.,    1973.     Reproductive  cycles   in   Oregon   populations   of  the   echinoid,    Strongylo- 

ccntrotus   purpuratus    (Stimpson).      II.     Seasonal    changes    in    oocyte   growth   and    in 

abundance  of  gametogenic  stages  in  the  ovary.    /.  E.vp.  Mar.  Biol.  Ecol.,  12:  65-78. 
GREEX,  J.   D.,   1976.     Seasonal  reproductive  biology  of  Leptosynapta   tennis    (Echinodermata: 

Holothuroidea).     M.  A.  Thesis,  University  of  North  Carolina,  Chapel  Hill,  128  pp. 
HOLLAND,   N.   D.,   J.   C.   GRIMMER,   AND  H.   KUBOTA,    1975.     Gonadal   development   during   the 

annual   reproductive  cycle  of  Comanthus   iapunica    (Echinodermata:   Crinoidea).   Binl. 

Bull,  148:219-242. 

HUMASON,  G.  L.,  1962.    Animal  tissue  techniques.    W.  H.  Freeman,  San  Francisco,  468  pp. 
HYMAN,  L.  H.,  1955.     The  Invertebrates.  }'ol.  IV.,  Echinodermata.    McGraw-Hill,  New  York, 

763  pp. 
KRISHNASWAMY,    S.,   AND   S.  KRISHNAN,   1967.     A   report   on  the  reproductive   cycle   of  the 

holothurian,  Holothuria  scabra  Jager.     Curr.  Sci.,  36:   155-156. 

KUME,  M.,  AND  K.  DAN,  1968.    Invertebrate  embryology.     Nolit,  Belgrade,  Yugoslavia,   (trans- 
lated from  the  Japanese  by  Jean  C.  Dan ;  originally  published  in  1957  by  Bai  Fu  Kan 

Press,  Tokyo),  605  pp. 
McCRARY,   A.   B.,    1969.      Zooplankton   in   Wrightsville    Sound.     Ph.D.    Thesis,    University    of 

North  Carolina,  Chapel  Hill,  135  pp.     (Diss.  Abstr.,  31:  63-B ;  order  no.  70-12,081.) 
MENKER,  D.,   1970.     Lebenszyklus,  Jugendentwicklung  und  Geschlechtsorgane  von  Rhabdomol- 

gus  rubcr  (Holothuroidea:    Apoda).    Mar.  Biol.,  6:  167-186. 
RUTHERFORD,  J.   C.,   1973.     Reproduction,   growth  and  mortality  of  the  holothurian   Cucitmariii 

pscudocurata.  Mar.  Biol. ,22:  167-176. 
STEEDMAN,   H.  F.,   1960.     Section  cuttinq  in   microscopy.     Blackwell   Scientific   Publ.,   Oxford. 

172  pp. 
TANAKA,  Y.,  1958.     Seasonal  changes  occurring  in  the  gonad  of  Stielmpus  iapnnicns.    Bull.  Fac. 

Fish.  Hokkaido  Univ.,  9:  29-36. 
TURNER,    V.   G.,    1966.      The   reproductive    biology   of   selected   echinoderms    from    Cape    Cod, 

Massachusetts.     Master's  Thesis,  University  of  California,  Los  Angeles,  137  pp. 
YOSHIDA,  M.,  1952.     Some  observations  on  the  maturation  of  the  sea  urchin,  }>iadcina  setosuin. 

Annot.  Zool.  Jpti.,  25:  265-271. 


Reference:  Biol.  Bull,  154:  79-95.      (February,   1978) 


DEVELOPMENT  OF  AMPHIOPLUS  ABDITUS  (VERR1LL*)   (ECHINO- 

DERMATA:  OPHIUROIDEA).    II.  DESCRIPTION  AND  DISCUSSION 

OF  OPHIUROID  SKELETAL  ONTOGENY  AND  HOMOLOGIES  x 

GORDON  HENDLER 2 
Smithsonian  Tropical  Research  Institute,  P.O.  Box  2072,  Balboa,  Canal  Zone 

The  descriptive  studies  of  the  last  century,  meticulous  and  seemingly  irrefutable, 
sometimes  conceal  serious  flaws.  The  comprehensive  scheme  of  ophiuroid  skeletal 
anatomy  of  Ludwig  (1878,  1881,  1899,  1901),  for  example,  is  presented  in  numer- 
ous treatises  (e.g.,  Bather,  Gregory,  and  Goodrich,  1900;  MacBride,  1906;  Cuenot, 
1948;  Hyman,  1955;  Spencer  and  \Yright,  1966)  as  a  cornerstone  for  theories  of 
echinoderm  systematics  and  phylogeny.  In  the  present  paper,  Ludwig's  deductions 
and  observations  on  ophiuroids  are  compared  and  contrasted  with  aspects  of  the 
anatomy  and  ontogeny  of  Ainphioplus  abditns,  and  a  new  interpretation  of  the 
ophiuroid  skeleton  is  suggested.  In  addition,  it  is  shown  that  juveniles  of  A.  abditns 
undergo  such  drastic  changes  in  skeletal  morphology  during  ontogeny,  that  dif- 
ferent developmental  stages  might  easily  be  mistaken  for  the  young  of  other  genera. 
The  implications  of  these  transformations  for  systematics  of  the  amphiurids  are 
considered.  Terminology  employed  for  this  treatment  is  based  on  accepted  names 
of  structures  in  the  adult  ophiuroid,  except  for  "buccal  scale"  (cf.,  Hyman,  1955; 
Thomas,  1962).  The  nomenclature  and  abbreviations  are  summarized  in  Figure  1. 

MATERIALS  AND  METHODS 

Amphioplus  abditns  is  a  burrowing  amphiurid  with  direct  development.  Its 
demersal,  lecithotrophic  embryos  develop  within  a  fertilization  membrane  that  rests 
on  the  sediment  (Hendler,  1977).  Specimens  were  collected  and  treated  as  pre- 
viously described  (Hendler,  1977).  Postlarval  stages  were  reared  in  vessels  of 
sediment,  partly  immersed  in  a  running  seawater  system  to  maintain  the  cultures 
near  field  temperatures.  Juveniles  were  collected  from  Beebe  Cove,  Noank,  Con- 
necticut, and  Wild  Harbor,  West  Falmouth,  Massachusetts. 

Skeletal  development  was  studied  using  whole  or  dissected  specimens  which 
were  dehydrated,  cleared,  and  mounted  in  "Permount."  The  hard-parts  were 
examined  with  both  regular  illumination  and  polarized  light,  and  Rose  Bengal  or 
Grenadier's  Borax  Carmine  were  used  to  stain  the  water-vascular  system  (Huma- 
son,  1967). 

RESULTS 
Larval  skeleton 

A  triangular  granule  appears  in  both  posterior  angles  of  the  24-hour  embryo, 
and  each  granule  grows  to  a  tetraradiate  spicule  within  five  hours.  By  33  hours, 

1  University  of  Connecticut  Marine  Research  Laboratory  Contribution  No.  106. 
-  Present   address :     Smithsonian    Oceanographic    Sorting    Center,    Smithsonian    Institution, 
Washington,  D.C.  20560. 

79 


80 


GORDON  HENDLER 


3r 
O     6>      <s) 

A,<T  ^  T 


FIGURE  1.  A.  Diagrammatic  portion  of  an  ophiuroid  jaw  frame,  in  dorsal  view  with  disc 
removed;  mouth  opening  is  to  the  left.  The  base  of  one  arm  and  two  half-jaws  are  illustrated; 
ambulacra  of  the  vertebrae  are  separated  by  dashes;  plates  of  the  jaw  shown  as  unfused  pieces. 
Stippled  area  —  oral  plate  =  half-jaw  =  proximal  oral  plate  +  distal  oral  plate.  B.  Diagrammatic 
three-dimensional  representation  of  an  ophiuroid  oral  frame  section  with  disc  removed ;  mouth 
opening  is  to  the  left.  An  entire  jaw  and  proximal  bases  of  two  arms  are  shown.  Ambulacra 
of  each  jaw  and  vertebra  are  shown  as  unfused  plates.  The  proximal  and  distal  oral  plates 
of  the  jaw  are  modified  and  abradially  directed  arm-vertebrae  (ambulacra)  ;  ambulacral  plates 
from  tii'o  arms  comprise  each  jaw.  AS  represents  adoral  shield  (adambulacral-2 )  ;  BS,  buccal 
scale;  BTS,  first  buccal  tube-foot  scale;  DAP,  dorsal  arm-plate;  DOP,  distal  oral  plate 
(ambulacral-2)  ;  DP,  dental  plate;  LAP,  lateral  arm-plate  (adambulacral)  ;  OS,  oral  shield; 
PER,  peristomial  plate;  POP,  proximal  oral  plate  (ambulacral-1)  ;  SP,  arm  spine;  T,  tooth; 
TF,  buccal  tube-foot;  TP,  terminal  plate;  VAP,  ventral  arm-plate;  and  VERT,  vertebra 
(ambulacrum) . 

the  four  points  of  the  spicule  lengthen  and  give  rise  to  lateral  and  terminal  branches 
while  several  short,  branching  extensions  accumulate  at  the  nexus  of  the  primary 


OPHIUROID  SKELETAL  DEVELOPMENT 


81 


rods  (Fig.  2).    Resorption  of  the  larval  skeleton  normally  begins  around  57  hours 
(Fig.  2),  and  between  74  and  84  hours,  as  the  embryonic  arms  disappear,  the 


B 


L 


MfrSg— BS 


OOP 


FIGURE  2.  Diagrams  showing  approximate  arrangement  and  relative  size  of  developing 
skeletal  elements  (in  ventral  view).  C  and  D  show  only  one  radial  portion  of  the  embryo: 
A,  late  triangular  embryo,  35  hr ;  B,  early  star-disc  stage,  55  hr ;  C,  late  star-disc  stage,  73  hr  ; 
and  D,  newly  hatched  juvenile,  96  hr.  Scale  line  is  approximately  0.1  mm.  Abbreviations  as  in 
Figure  1,  and  ASP  represents  adoral  shield  spine;  C,  central  plate;  G,  granules  of  presumptive 
spicules;  IR,  interradial-1  ;  L,  larval  skeleton;  R,  radial  plate;  and  TP,  terminal  plate. 


82 


GORDON  HENDLKR 


TAHLE  I 
Chronology  of  skeletal  element  formation. 


Disc 

Time 

diameter 

Aboral  surface 

Oral  surface 

(mm) 

35  hr 

Radial  plates,  central 

plates,  terminal  plates 

35-42  hr 

Adoral  shields,  proximal  oral 

plates  (POP) 

45-55  hr 

Distal  oral  plates  (OOP) 

48  hr 

Interradial  1  (initial) 

74  hr 

Adoral  shield  spines 

84  hr 

Dental  plates 

96-147  hr 

Interradial-1  (2nd-5th) 

117  hr 

Ventral  arm-plate-  1  (VAP) 

110-147  hr 

0.3 

Teeth 

55  days 

Lateral  arm-plate-  1   (LAI') 

55-72  days 

Vertebrae,  lateral  arm-plate 

spines,  ventral  arm-plate-2 

Number  of  arm 

segments 

2-3  (<5  months) 

0.4 

Dorsal  arm-plate-1   (DAP), 

madreporite,  oral  shields 

3-4  (8  months) 

Radial  shields 

5 

0.7 

1  nterradial-2 

Infradental  papillae,  ventral  disc- 

scales 

7 

Peristomials 

Second  teeth 

8 

First  buccal  tube  foot  scales 

9 

0.9 

Genital  scales 

15 

Genital  plates 

17 

1.1 

Third  oral  papillae,  fourth  oral 

papillae  (from  adoral  shield 

spines) 

20 

1.3 

>30 

1.6 

»30 

1.9 

Interradial  muscle  scales, 

Tentacle  scales,  accessory  scales 

dorsal  disc  scales 

>41 

2.3 

»38 

2.5 

82 

3.5 

branched  skeletons  dwindle  to  straight  pieces  \vith  furcate  tips  and  then  are  lost 
(Fig.  2). 

The  larval  skeleton  of  Atnphioplus  compares  with  that  of  the  typical  ophio- 
pltiteus,  the  four  major  branches  corresponding  to  the  body,  posterolateral,  antero- 
lateral  and  postoral  rods  of  planktonic  ophiuroid  larvae.  The  body  rods  of  A. 
abditus  bear  branched  tips  which  may  lie  homologous  to  transverse  and  end  rods, 
even  though  the  tips  do  not  articulate.  However,  the  general  shape  and  pattern  of 
secondary  branching  of  the  skeleton  varies  between  specimens,  and  for  each  individ- 
ual the  skeleton  in  one  arm  is  larger  or  more  complex  than  its  counterpart  in  the 
other  arm. 

It  is  noteworthy  that  by  35  hours  of  development  separate  portions  of  the  larval 


OPHIUROID  SKELETAL  DEVELOPMENT 


83 


skeleton,  but  not  the  ophiuroid  skeletal  elements,  have  distinct  and  different  angles 
of  extinction  under  polarized  light.  In  other  words,  a  larval  skeletal  element, 
which  grows  from  a  tetraradiate  spicule,  does  not  act  as  a  single  calcite  crystal  hut 
is  composed  of  irregular  segments  with  different  crystallographic  orientations.  This 
condition,  unusual  for  echinoderm  skeletons,  may  he  an  effect  of  a  peculiarity  in  the 
coordination  between  the  skeleton-depositing  cells. 

Ophiuroid  skeleton 

Between  30  and  35  hours,  the  radial,  central,  and  terminal  plates  appear  and 
regroup  from  a  sagittal  plane  to  form  concentric  rings  on  the  ventral  surface  of  the 
embryo  (Fig.  2,  Table  I).  Judging  from  the  relative  size  (breadth  rather  than 
mass)  of  the  spicules  in  slightly  older  specimens,  the  radial,  central  and  terminal 
plates  are  produced  in  that  order. 

Minute,  granular  rudiments  of  the  proximal  oral  plates  and  adoral  shields  are 
visible  by  35  hours  (Fig.  2).  By  42  hours  the  adoral  shield  rudiments  are  larger 
than  the  proximal  oral-plate  rudiments,  and  the  spicule  is  triradiate.  Some  45-hour 
specimens  have  granular  rudiments  of  the  buccal  scales  and  the  distal  oral  plates. 

By  55  hours  the  embryonic  ophiuroid  rudiment  is  pentaradiate,  and  its  skeletal 
plates  are  arranged  as  successive  stacks  of  concentric  rings.  In  the  dorsal-most 
plane  are  central  and  radial  plates ;  terminal  plates,  shaped  like  stick  figures  with 
outstretched  arms,  are  more  ventral  and  on  the  perimeter  of  the  disc;  the  branched 
proximal  oral  plates  and  adoral  shields  are,  respectively,  proximal  and  distal  to  the 
center  of  the  disc  and  at  a  level  below  the  terminals;  and  buccal  scale  granules  and 
triradiate  distal  oral  plates  are  near  the  ventral  surface  (Fig.  2). 

In  some  specimens,  after  45-55  hours  of  development,  a  single  triradiate  spicule, 
a  rudiment  of  the  initial  aboral  interradial  plate  (interradial-1),  appears  at  the  edge 
of  the  ophiuroid  rudiment,  ventral  to  the  terminal  plates  and  at  the  right  side  of 
the  embryo  in  the  interradius  anterior  to  the  right  larval  skeleton.  Interradial-1 
plates  do  not  appear  in  the  other  four  interradii  until  96  hours  of  development. 

A  B  C 


FIGURE  3.  Portion  of  the  disc,  in  ventral  view,  for  different  stages  of  development:  A,  1-2 
arm-segment  stage;  B,  3  arm-segment  stage;  and  C,  5  arm-segment  stage.  Abbreviations  as  in 
Figures  1  and  2;  and  IP  represents  infradental  papilla;  R,  radial  plate.  Scale  line  is  0.1  mm. 
Dotted  lines  separate  DOP  and  POP  as  seen  in  polarized  light. 


84 


GORDON  HENDLER 


Triradiate  rudiments  of  the  adoral  shield  spines  appear  at  74  hours,  and  by  84  hours 
the  dental  plates  materialize  as  spicules  proximal  to  the  five  jaws  (Figs.  2  and  3). 
From  90  to  110  hours  the  skeletal  elements  of  hatching  juveniles  become  denser 
and  approach  their  final  form  (Figs.  2  and  3). 

By  96  hours,  there  are  triradiate  spicules,  rudiments  of  the  first  interradial 
plates,  in  a  plane  beneath  the  radial  plates  (except  for  the  radius  with  a  precocious 
interradial-1).  By  147  hours,  these  spicules  form  small  plates  with  multiple 
branches ;  by  23  days  they  are  larger  and  nearly  perpendicular  to  the  radial  plates, 
and  they  then  migrate  to  the  dorsal  surface  of  the  disc. 

The  adoral  shield  spines  move  to  the  shield  and  grow  beyond  the  edge  of  the 
disc.  At  the  time  of  hatching,  the  juvenile  has  its  dorsal  surface  shingled  by  a 
rosette  of  overlapping  primary  plates  and  around  the  disc  there  are  spike-like,  pro- 
truding terminal  plates  and  adoral  shield  spines.  Shortly  before  hatching  (about 
167  hours)  on  the  ventral  surface  of  the  disc,  the  distal  end  of  the  proximal  oral 
plate  enlarges  and  fuses  with  the  distal  oral  plate  (the  net-like  stereom  of  the 
latter  remains  distinctive  even  after  fusion).  Pairs  of  oral  plates  articulate  to  form 
jaws  by  230  hours. 

Several  days  after  hatching  (by  196  hours)  the  first  series  of  teeth,  initially 
sparsely-branched  triradiate  spicules,  form  at  the  proximal  surface  of  the  dental 
plate,  but  they  do  not  take  their  final  massive,  blunt-tipped  form  until  the  9  to  17 
arm-segment  stage  of  development. 

Rudiments  of  the  ventral  arm-plates,  already  present  at  117  hours,  become 
dense,  fenestrate  plates  with  an  elongate  pentagonal  shape  by  13  days.  Opacity  of 
the  ventral  arm-plates  and  the  appearance  of  lateral  arm-plates  by  55  days  obstruct 
examination  of  the  vertebrae,  although  slender,  unfused  vertebral  ossicles  are  still 
discernable  until  72  days  when  the  first  arm  spines  and  second  series  of  ventral  arm- 
plates  appear.  The  first  dorsal  arm-plates  are  added  to  the  arm  before  three  months. 


B 


FIGURE  4.  Portion  of  the  disc,,  in  ventral  view,  for  different  stages  of  development:  A,  9 
arm-segment  stage;  and  B,  17  arm-segment  stage.  Abbreviations  as  in  Figures  1  to  3 ;  and 
3  represents  third  oral  papilla;  VDS,  ventral  disc  scales.  Scale  line  is  0.1  mm.  Dotted  lines 
separate  DOP  and  POP. 


OPHIUROID  SKELETAL  DEVELOPMENT 


B 


85 


PER 


FIGURE  5.  Portion  of  the  disc,  in  ventral  view,  for  different  stages  of  development:  A,  21 
arm-segment  stage;  and  B,  >30  arm-segment  stage.  Abbreviations  as  in  Figures  1  to  4 ;  and 
AP  represents  accessory  papillae;  GS,  genital  scale;  TS,  tentacle  scale.  Scale  line  is  0.1  mm. 
Dotted  lines  separate  DOP  and  POP. 

As  each  new  arm  segment  is  added  a  ventral  arm-plate  appears  before  the  lateral 
arm-plates  and  arm  spines,  and  a  dorsal  arm-plate  finally  forms  to  cap  the  segment. 
The  arm-plates  and  spines  take  a  form  characteristic  of  the  adult  by  the  17  arm- 
segment  stage. 

Rudiments  of  the  oral  shields  and  madreporite  do  not  appear  until  the  third 
arm-segment  begins  to  form  (by  five  months).  They  form  in  situ  on  the  oral 
surface  of  the  disc,  proximal  to  the  interradial-1  series  (Fig.  3).  Even  so,  the 
stone  and  pore  canals  associated  with  the  madreporite  are  present  before  the  two 
arm-segment  stage.  By  eight  months  there  are  four  arm-segments,  and  radial 
shields  appear  at  the  base  of  the  arm,  on  the  ventral  surface  of  the  disc. 

An  exact  chronology  cannot  be  assigned  for  subsequent  skeletal  developments, 
as  they  were  deduced  using  a  growth  series  organized  from  material  in  field  collec- 
tions (Table  I;  Figs.  3,  4,  and  5).  There  is  no  important  change  in  the  armature 
of  the  disc  until  the  5  arm-segment  stage,  when  pairs  of  infradental  papillae  appear 
distal  to  the  dental  plates  (Fig.  3).  The  most  striking  superficial  change  at  this 
time  is  the  addition  of  minute  scales  on  the  ventral  surface  of  the  disc,  which 
separate  the  oral  shields  and  interradial-1  plates.  Interradial-2  plates  erupt  be- 
tween the  radials,  isolating  the  radial  plates  from  the  interradial-1  on  the  dorsal 
surface  of  the  disc  (Figs.  3,  4,  and  5).  The  central  primary  plate  and  the  radial 
plates  continue  to  grow  until  the  disc  diameter  is  5  mm  and  then  diminish  in  size 
with  further  growth.  This  allometric  pattern  is  species-specific  (Hendler,  1973). 

There  are  three  important  series  of  minute  plates  within  the  disc.  Reticulate 
peristomial  plates  are  added  by  the  7  arm-segment  stage  on  the  inner  surface  of  the 
jaw  between  the  proximal  and  distal  oral  plates  (Fig.  6).  They  probably  arise  as 
single  spicules  at  an  earlier  stage,  but  there  is  no  doubt  that  the  peristomial  plates 
and  buccal  scales  are  present  at  the  same  time  (Fig.  6).  By  the  8  arm-segment 
stage,  pairs  of  tiny  plates  form  within  the  disc,  distal  to  the  buccal  scales,  and  just 
proximal  to  each  first  ventral  arm-plate.  These  plates  will  be  referred  to  below  as 


86 


GORDON  HENDLER 


B 


FIGURE  6.  Portion  of  the  oral  frame,  in  dorsal  view,  with  disc  removed,  for  different 
stages  of  development:  A,  7  arm-segment  stage;  B,  13  arm-segment  stage;  C,  >30  arm-segment 
stage ;  and  D,  >41  arm-segment  stage.  Abbreviations  as  in  Figures  1  to  4 ;  and  GP  represents 
genital  plate;  GS,  genital  scale;  MS,  interradial  muscle  scale.  Scale  line  is  0.1  mm.  Note  that 
buccal  scales,  which  lie  on  the  oral  surface,  cannot  be  seen  in  C  or  D.  Dotted  lines  separate 
DOP  and  POP. 


the  first  buccal  tube-foot  scales  (corresponding  plates  have  been  treated  as  ventral 
arm-plates  or  homologues  of  anilmlacral  plates  by  various  authors). 

By  the  9  arm-segment  stage,  due  to  growth  of  the  adoral  shield,  the  adoral 
shield  spine  sits  near  the  center  of  the  shield  plate  (Fig.  4).  It  then  acts  as  a 
"tentacle  scale"  for  the  second  buccal  tube-feet,  analogous  to  the  minute  plates  on 


OPHIUROID  SKELETAL  DEVELOPMENT  87 

the  ventral  surface  of  the  arm  that  shield  the  contracted  tube-feet  (tentacles).  The 
buccal  scales  transform  from  crescentric  scales  to  blunt  spines.  The  madreporite 
exhibits  an  external  perforation  for  the  hydropore.  The  genital  scales  were  seen  in 
a  9  arm-segment  individual;  and  genital  plates  were  seen  in  a  specimen  of  15  arm- 
segments,  but,  undoubtedly,  both  originated  at  an  earlier  stage. 

By  the  17  arm-segment  stage  the  rudiments  of  the  third  oral  papillae  appear  on 
the  distal  surface  of  the  oral  plates  (Fig.  4).  The  spines  of  the  adoral  shield  mi- 
grate to  form  additional  oral  papillae  (the  fourth  oral  papillae).  Thus,  by  the  21 
arm-segment  stage,  each  row  of  oral  papillae  consists  of  a  block-like  infradental 
followed  by  a  spine-like  papilla  (formed  by  a  buccal  scale),  a  third  papilla  (chrono- 
logically the  youngest),  and  a  large  distal  papilla  (formed  by  an  adoral  shield  spine) 
(Fig.  5).  Oral  shields  of  adjacent  jaws  are  initially  tightly  appressed,  but  the  first 
ventral  arm-plates,  which  begin  migrating  into  the  buccal  area  by  the  5  arm-seg- 
ment stage,  spread  apart  the  adoral  shields. 

When  more  than  30  arm-segments  have  formed,  the  oral  plates  begin  to  take  a 
definitive  shape  with  enlarged  lateral  extensions  for  the  insertion  of  the  external 
interradial  muscles  and  a  deep  cleft  for  the  water-vascular  ring  (Fig.  6).  In  addi- 
tion, a  scattering  of  small  thin  plates  accumulates  on  the  surface  of  the  interradial 
muscles  (Fig.  6).  The  paired  peristomial  plates  lengthen  and  then  degenerate  (or 
even  fuse?),  becoming  fragile  scales  like  those  on  the  interradial  muscles.  The 
rudimentary  skeletal  elements  proximal  to  the  ventral  arm-plates  (first  buccal 
tube-foot  scales)  form  reticulate  pieces  perpendicular  to  the  axis  of  the  arm.  By  this 
time  tentacle  scales  have  appeared  on  many  arm-segments,  one  scale  per  tube-foot. 

The  first  buccal  tube-foot  scales  located  within  the  oral  cavity  surpass  the  size 
of  the  peristomial  plates ;  in  larger  specimens  they  form  thin,  boomerang-shaped 
elements  beside  the  inner  oral  papillae.  While  the  peristomial  plates  stop  growing 
and/or  degenerate,  the  scales  over  the  interradial  muscles  grow  larger  and  more 
numerous,  obscuring  the  peristomials.  The  interradial  muscle  scales  finally  form  a 
medial  band  on  the  muscles  in  the  adult  (especially  dense  on  the  muscle  above  the 
madreporite)  (Fig.  6). 

On  the  dorsal  surface  of  the  disc,  the  number  of  tiny  scales  separating  the  pri- 
mary and  interradial  plate  continues  to  grow,  but  two  series  of  interradial  plates 
remain  distinct.  The  interradial- 1  plates  remain  on  the  edge  of  the  disc  until  the 
specimens  have  at  least  82  arm-segments  (disc  diameter  =  3.5  mm),  and  in  larger 
animals  they  are  left  on  the  dorsal  surface  of  the  disc  as  the  rate  of  disc  growth 
increases  centrifugally.  On  the  ventral  surface  of  the  disc,  small  accessory  scales 
(the  fifth  oral  papillae)  form  distally  to  the  fourth  oral  papillae,  and  additional 
accessory  scales  may  appear  in  later  stages  of  growth  (Fig.  5). 

DISCUSSION 

The  derivation  of  the  oral  frame  and  the  armament  of  oral  papillae  bordering 
the  buccal  cavity  are  discussed  in  detail,  as  these  are  the  features  of  A.  abditus  that 
best  illustrate  the  systematics  of  amphiurids  (the  largest  family  of  ophiuroids)  and 
the  affinities  of  the  echinoderm  classes.  The  apical  rosette  of  primary  plates,  radial 
shields,  and  the  madreporite  will  be  discussed  first,  since  these  structures  figure 
prominently  in  evaluations  of  echinoderm  phylogeny. 


88  GORDON  HENDLER 

In  the  ophiuroids  the  larval  skeleton  plays  no  part  in  the  formation  of  the  apical 
system  (central  and  radial  primary  plates),  since  the  larval  skeleton  of  ophiuroids 
is  resorbed,  as  in  .•/.  abditus,  or  sometimes  "discarded"  (Mortensen,  1931  ;  Olsen, 
1942;  Hendler,  1975).  In  echinoids,  on  the  other  hand,  the  larval  skeletal  pieces 
generate  the  genital  and  terminal  apical  plates  (Onoda,  1931).  The  origin  of  both 
the  larval  skeleton  and  apical  plates  has  been  thought  to  provide  evidence  for  a 
close  affinity  between  the  echinoids  and  ophiuroids.  However,  the  divergence  seen 
in  the  ontogeny  of  ophiuroids  and  echinoids  implies  that  there  is  nonhomology  of 
the  larval  skeleton  and/or  the  apical  plates  of  the  two  classes. 

Regardless  of  discrepancies  in  homology  between  ophiuroids  and  echinoids,  the 
central  plate  of  ophiuroids  appears  to  be  homologous  to  the  central  plate  of  aster- 
oids. Apart  from  this  characteristic,  the  similarities  between  the  discs  of  ophiuroids 
and  asteroids  are  few.  Prominent  structures  of  the  ophiuroid  disc,  such  as  radial 
shields,  oral  shields,  and  genital  plates  and  scales,  seem  to  be  ophiuroid  specializa- 
tions without  homologues  in  asteroids. 

Each  radial  shield  (and  the  other  major  skeletal  elements)  of  A.  abditus  and 
other  ophiuroids  originates  from  a  single  element  (not  by  fusion  of  scales  as  sug- 
gested in  Spencer  and  Wright,  1966).  The  radial  shields  (and  genital  plates  and 
scales)  reappear  when  the  disc  of  Amphioplus  regenerates,  but  the  primary  plates 
(loci  of  disc  scale  formation)  never  are  replaced  (personal  observation).  This 
morphogenetic  difference  between  the  central  plate  and  radial  shields  indicates  that 
the  primary  plates  of  ophiuroids  are  atavistic  structures  that  are  recapitulated  dur- 
ing ontogeny  but  are  not  necessary  for  the  growth  of  the  disc. 

Even  structures,  such  as  the  madreporite,  which  have  been  judged  to  show  close 
affinities  between  the  ophiuroids  and  asteroids,  point  up  differences  between  the 
classes.  Although  in  the  recent  ophiuroids  an  oral  shield  acts  as  the  madreporite, 
some  primitive  ophiuroids  possess  a  madreporite  but  lack  oral  shields  (Spencer  and 
Wright,  1966).  It  is  possible  that  a  madreporite  formed  of  an  oral  shield  is  an 
innovation  in  the  recent  ophiuroids,  but  a  different  structure  held  the  water-pore  in 
the  ancient  ophiuroids.  In  other  words,  the  madreporites  may  be  nonhomologous 
in  recent  ophiuroids,  ancient  ophiuroids,  and  asteroids.  Furthermore,  the  "preco- 
cious interradial"  plate  and  early  formation  of  the  hydroporic  canal  observed  in 
A.  abditus  could  be  vestiges  of  the  primitive  madreporite. 

The  madreporite  of  asteroids  is  usually  dorsal,  and  ophiuroids  generally  have  a 
ventral  madreporite.  Hence,  the  relationship  of  these  two  classes  has  been  gauged 
by  comparing  the  point  of  origin  of  the  ophiuran  madreporite :  whether  on  the 
ventral  hemisome  (indicating  lack  of  affinity  with  asteroids),  or  by  way  of  migra- 
tion from  the  dorsal  surface  (indicating  a  close  relationship)  (Ludwig,  1881 ; 
Mortensen,  1912,  1921;  Murakami,  1940,  1941).  In  fact,  all  recent  ophiuroids  re- 
ported to  have  a  ventrally  migrating  madreporite  may  actually  have  a  "precocious 
interradial."  For  example,  Murakami  (1940)  described  a  dorsal  to  ventral  mi- 
gration of  the  madreporite  in  Axiognathus  (  —  Amphipholis)  squanmtus,  but  his 
illustrations  show  that  the  "madreporite"  is  identical  to  the  "precocious  interradial" 
of  A.  abditus;  so  the  true  madreporite  of  Axiognathus  must  originate  at  its  final, 
ventral  location.  This  suggests  that  formation  of  the  madreporite  in  situ  on  the 
ventral  surface  is  a  characteristic  distinguishing  recent  ophiuroids  from  asteroids. 
But,  without  more  information,  the  precocious  interradial  of  ophiuroids  cannot  be 


OPH1URO1I)   SKKLKTAL  DEVELOPM  FAT  89 

considered  a  homologue  of  the  interradial  plate,  which  acts  as  a  madreporite,  in 
recent  asteroids. 

Just  as  the  superficial  structures  of  the  disc  discussed  above,  such  as  the  apical 
plates,  radial  shields  and  madreporite,  suggest  drastic  distinctions  between  classes 
of  echinoderms,  basic  differences  in  the  formation  of  the  oral  frame  further  em- 
phasize the  divergence  between  ophiuroids  and  asteroids.  In  the  ophiuroid  embryo 
the  oral  frame  is  produced  by  a  rearrangement  of  the  paired,  serial,  skeletal  elements 
that  fuse  to  form  "vertebrae"  inside  each  arm.  At  the  base  of  the  arm,  instead 
of  forming  vertebrae,  the  elements  fuse  with  corresponding  pieces  in  the  neighbor- 
ing arm,  making  a  connecting  bridge  between  adjacent  arms  that  projects  into  the 
oral  cavity  as  a  tooth-bearing  jaw.  Half  of  each  bridge  is  an  "oral  plate"  com- 
prised of  elements  from  a  single  arm.  Each  jaw  consists,  in  effect,  of  two  oral  plates 
(half-jaws)  and  accompanying  elements  from  two  contributing  arms. 

Jn  his  studies  of  this  "oral  frame  system,  Ludwig  (1878,  1881,  1889,  1901) 
homologized  the  ophiuroid  and  asteroid  oral  frames.  He  considered  the  proximal 
oral  plates  of  the  ophiuroid  jaw  (the  proximal  portion  of  each  half- jaw)  to  be  the 
first  adambulacral  plates,  the  adoral  shields  as  the  second  adambulacrals,  and  the 
lateral  arm-plates  as  serially  analogous  adambulacrals.  He  described  the  fusion  of 
the  proximal  and  distal  oral  plates  during  ontogeny  and  claimed  the  distal  oral  plates 
to  be  the  second  ambulacrals,  serially  homologous  to  the  vertebrae  of  the  arm.  Lud- 
wig believed  the  peristomial  plates  developed  from  the  early-forming  buccal  scales 
and  felt  they  represented  the  first  ambulacrals  of  the  jaw.  This  relationship  be- 
tween the  peristomial  plates  and  the  buccal  scales,  and  their  identification  as  the 
first  ambulacral  plates,  are  both  erroneous.  Consequently,  Ludwig's  analogies  be- 
tween ophiuroid  and  asteroid  jaws  are  not  legitimate. 

Zur  Strassen  (1901)  pointed  out  that  the  peristomials  (0  to  3  in  number  in 
different  species)  were  not  paired  in  each  jaw  as  proper  ambulacrals  must  be. 
Moreover,  he  observed  that  buccal  scales,  which  Ludwig  proposed  to  be  the  rudi- 
ments of  peristomial  plates,  were  present  concurrently  with,  and  hence  unrelated 
to,  peristomial  plates.  In  the  species  that  zur  Strassen  studied,  the  peristomials 
persist  and  the  buccal  scales  are  resorbed.  In  contrast,  in  A.  abditus  the  peristomial 
plates  are  lost  and  the  buccal  scales  become  oral  papillae.  These  observations  indi- 
cate a  basic  flaw  in  Ludwig's  scheme,  calling  into  question  the  identity  of  the  first 
plate  of  the  ambulacral  series  and  the  affinities  of  the  oral  plate  elements  and  buccal 
scales. 

Just  as  Ludwig  proposed,  the  adoral  shield  is  adambulacral  and  serially  ho- 
mologous with  the  lateral  arm-plates,  as  indicated  by  its  position  in  relation  to  the 
arm  and  its  transitory  possession  of  a  spine  (Simroth,  1876;  Fewkes,  1887; 
Murakami,  1937).  The  distal  oral  plate,  clearly  associated  with  the  water-vascular 
system,  is  obviously  an  ambulacral  plate,  but  the  identity  of  the  proximal  oral  plate 
is  an  unresolved  problem.  Here  it  is  suggested  that  the  proximal  oral  plate,  rather 
than  the  peristomial,  is  the  first  ambulacral  plate  of  the  jaw. 

Alone,  the  association  of  two  pairs  of  tube-feet  \vith  the  jaw  indicates  that  two 
elements  of  the  oral  plate  are  ambulacral,  and  there  are  only  two  parts  of  the  jaw 
per  sc,  the  proximal  and  distal  oral  plates.  The  attachment  of  the  proximal  oral 
plate  in  a  series  with  the  distal  oral  plate  suggests  that  it  is  an  ambulacral  element, 
but  the  buccal  tube-feet  of  A.  abditus  penetrate  only  the  distal  oral  plate,  suggesting 


90  GORDON  HENDLRR 

that  the  proximal  oral  plate  is  not  amlmlacral.  But,  just  as  the  adoral  shields  are 
homologous  to  lateral  arm-plates,  although  the  orientations  of  the  shields  and  plates 
are  different,  the  obliquely  oriented  proximal  oral  plates  may  he  amhulacral  even 
though  they  are  bypassed  by  the  \vater-vascular  system.  The  separation  of  the 
proximal  oral  plate  and  the  water-vascular  system  may  simply  result  from  the 
formation  of  buccal  tentacles  in  reverse  order  in  A.  abditiis  and  other  ophiuroids 
(Miiller,  1851;  Krohn,  1851;  Apostolides,  1882;  Grave,  1900;  MacBride,  1906; 
Mortensen,  1921;  Fell,  1941,  1946)  and  the  inward  migration  of  the  buccal  scales 
and  marked  dorsal  rotation  of  the  ventral  arm-plates. 

Thus,  the  jaw  of  ophiuroids  must  constitute  two  transformed  arm-segments 
which  are  highly  modified  and  "incomplete."  In  the  first  "segment"  there  is  an 
ambulacral  proximal  oral  plate,  but  the  identity  of  the  adambulacral  plate  is  proble- 
matical, and  homologues  of  accessory  plates  are  lacking.  The  second  "segment" 
consists  of  the  distal  oral  plate  (ambulacral),  adoral  shield  (adambulacral),  and  the 
ventral  arm-plate.  There  seems  to  be  no  unequivocal  homologue  of  the  adambu- 
lacral-1  in  the  oral  frame  of  recent  ophiuroids.  As  explained  above,  the  peristomial 
plates  are  obviously  secondary  structures  (not  adambulacrals).  The  buccal  scales, 
however,  might  be  adambulacrals,  but  without  corroborating  evidence  they  cannot 
be  considered  a  homologue  of  the  adambulacral  plates  in  the  first  modified  arm- 
segment.  The  occurrence  in  primitive  ophiuroids  of  a  typical  jaw  structure,  and 
the  presence  of  two  pairs  of  oral  tentacles  and  a  modified  proximal  oral  plate 
(Sollas  and  Sollas,  1912;  Schuchert,  1915;  Spencer,  1951),  demonstrate  both  the 
ambulacral  nature  of  the  proximal  oral  plates  and  a  basic  class-wide  concord  in 
ophiuroid  jaw  morphology.  Besides  the  ambulacral  affinity  of  the  proximal  oral 
plate,  the  fossil  record  suggests  a  recent  origin  for  the  peristomial  plates  and  the 
buccal  scales,  as  oral  papillae  and  buccal  scales  appear  to  be  absent  in  ophiuroids 
prior  to  the  Silurian  (Spencer  and  Wright,  1966).  Such  negative  evidence  must 
be  weighed  against  the  high  probability  of  preservational  bias,  especially  when 
minute  elements  such  as  these  papillae  are  concerned. 

The  ambulacral  nature  of  the  proximal  oral  plate  makes  the  ophiuroid  jawr 
arrangement  consistent  with  that  in  the  "somasteroid"  taxa,  while  in  other  asteroids 
the  proximal  element  of  the  jaw  is  an  adambulacral  homologue  (Fell,  1963;  Turner 
and  Dearborn,  1972).  Besides  this  major  contrast  between  the  classes,  differences, 
such  as  the  fusion  of  different  ambulacrals — numbers  two  and  three  in  somasteroid 
forms  (Fell,  1963),  numbers  one  and  two  in  the  ophiuroids — and  no  fusion  in  the 
asteroids,  also  reflect  the  considerable  divergences  between  ophiuroids  and  aster- 
oids. Clearly,  in  light  of  the  differences  in  homology  of  the  oral  frame,  the  dif- 
ferences in  ontogeny  and  morphogenesis  of  the  madreporite  and  primary  and  inter- 
radial  plates,  and  the  unique  structures  of  ophiuroids  such  as  radial  shields  and 
genital  plates  and  scales,  the  relationship  of  asteroids  and  ophiuroids  begs  re- 
evaluation. 

The  examination  of  skeletal  ontogeny  has  revealed  discrepancies  in  the  phylog- 
eny  of  the  echinoderm  classes,  but  ontogeny  of  the  oral  armature  can  indicate 
systematic  relationships  within  the  family  Amphiuridae.  Therefore,  before  discuss- 
ing the  systematics  of  the  amphiurids,  the  origins  of  the  oral  papillae  are  reviewed. 

In  A.  abditus  the  infradental  and  third  oral  papillae  and  the  fifth  (accessory) 
papillae  originate  in  situ.  They  are  secondary  structures  (i.e.,  not  of  the  ambu- 


OPHIUROID  SKELETAL  DEVELOPMENT  9t 

lacral-adambulacral  skeletal  groundwork).  In  contrast,  the  second  and  fourth 
papillae  migrate  to  their  ultimate  positions  while  undergoing  a  transformation  of 
shape  and  function. 

The  fourth  (distal)  papillae  originate  as  formidable  spines  on  the  acloral  shield 
and  are  used  initially  for  locomotion  and  balance,  but  they  are  dwarfed  and  re- 
located during  growth  of  the  oral  frame  and  ultimately  are  retained  as  diminutive 
scales  on  the  oral  plate.  Adoral  shield  spines  of  other  species  develop  as  in  A. 
abditus,  disappear  during  development,  or  even  take  a  peripheral  position  on  the 
disc  (Mortensen,  1933b;  Murakami,  1941;  Schoener,  1967). 

In  A.  abditus,  the  only  species  of  ophiuroid  whose  buccal  scales  have  been 
traced  through  development  to  the  adult  stage,  the  buccal  scales  of  the  juvenile 
are  among  the  first-formed  and  most  prominent  skeletal  elements.  It  is  remarkable 
that  they  initially  close  the  oral  gap  but  as  their  growth  slows  in  relation  to  that  of 
the  oral  plate,  they  "sink"  into  the  oral  slit  and  become  the  diminutive  second  oral 
papillae.  In  contrast,  the  buccal  scales  are  resorbed  during  development  in 
Axiognathus  (-- Amphipholis}  japonicus  and  A.viognathus  (-- Amphipholis} 
squamatus  (Ludwig,  1899;  /.ur  Strassen,  1901 ;  Sollas  and  Sollas,  1912;  Mortensen, 
1912,  1913,  1933a,  1938;  Murakami,  1937,  1940,  1941  ;  Guille,  1964).  Whether  all 
ophiuroid  species  have  buccal  scales  is  not  known. 

Since  the  second  oral  papillae  arise  from  the  buccal  scales  in  Amphioplus,  and 
they  differ  from  tentacle  scales  in  their  time  of  origin  and  mode  of  growth,  they 
should  not  be  called  "oral  tentacle  scales"  (Y/.,  A.  M.  Clark,  1970).  The  term 
"buccal  scales"  may  lie  used  for  the  early-forming  "spoon-shaped"  plates  of  the 
buccal  gap  whether  or  not  they  are  resorbed  in  the  adult ;  but  the  term  "second 
oral  papillae"  should  be  used  instead  of  "oral  tentacle  scales"  for  adult  amphiurids. 

It  is  not  certain  whether  both  sets  of  buccal  tube-feet  have  tentacle  scales.  The 
accessory  scales  (fifth  oral  papillae)  associated  with  the  second  buccal  tube-feet, 
which  are  obviously  tentacle  scales,  arise  long  after  the  other  four  oral  papillae  and, 
as  the  individual  grows,  they  increase  in  number  to  a  maximum  of  more  than  five 
per  tube-foot  (Hendler,  1973).  There  are  calcareous  elements,  proximal  to  the 
first  ventral  armplates  in  .1.  abditns  which  form  during  the  7-  to  9-arm-segment 
stage,  that  may  be  tentacle  scales  of  the  first  buccal  tube-feet.  They  resemble 
internal  buccal  elements  described  in  other  species  (Ludwig,  1878;  Mortensen, 
1912;  zur  Strassen,  1901;  Sollas  and  Sollas,  1912).  If  these  elements  are  not 
tentacle  scales,  then  the  first  buccal  tube-feet  lack  scales. 

From  the  early  appearance  of  infradental  and  distal  oral  papillae  in  A.  squamatus 
and  the  putative  juveniles  of  Amphioplus  acutus,  it  has  been  inferred  that  these 
species  pass  through  an  "Amphiura"  stage  (H.  L.  Clarke.  1914;  Mortensen,  1936). 
Such  phylogenetic  speculations  are  vulnerable,  being  based  on  a  single  character 
which  (in  the  case  of  Amphiura  oral  armature)  could  easily  be  paedomorphic  or 
secondarily  reduced.  Juveniles  of  A.riot/natlius  or  Amphioplus  may  resemble  early 
stages  of  Amphiura  or  even  Amphiodia,  but  it  cannot  be  assumed  that  these  genera 
recapitulate  an  "Amphiura  stage"  before  homologues  of  the  oral  papillae  of  all 
amphiurid  genera  are  delineated. 

On  the  basis  of  their  oral  armature,  the  amphiurids  may  be  divided  provisionally 
into  two  categories:  those,  like  Amphiura  and  Amphioplus,  with  the  second  oral 
papillae  located  high  on  the  oral  plate;  and  those,  like  Amphiodia  and  Amphipholis 


92  GORDON  HENDLER 

or  Axiognathus,  that  lack  second  oral  papillae.  It  is  shown  ahove  that  the  buccal 
scale  develops  into  the  second  papilla  in  ./.  ahditus.  However,  zur  Strassen  (1901), 
Murakami  (1940),  and  others  described  resorption  of  the  buccal  scale  in  Axio- 
(/iititlins  (=  Amphipholis)  species,  and  these  findings  have  been  confirmed  in 
Axiognathus  sqiiainatus  from  New  England  (personal  observation).  Thus,  it  is 
predicted  that  Amphiodia  and  Amphipholis  (like  Axiognathus)  resorb  the  buccal 
scales,  while  Am  phi  it  ra  species  (like  Amphioplus)  retain  the  scales  as  the  second 
oral  papillae. 

The  adoral  shield  spines  develop  in  different  ways  in  different  families  and 
genera,  but  it  would  be  interesting  to  see  whether  the  adoral  shield  spines  of 
amphiurids  always  become  distal  oral  papillae,  demonstrating  uniformity  in  the 
ontogeny  of  the  amphiurid  jaw.  The  adoral  shield  spine  of  Axiognathus  japonicus 
develops  identically  to  its  homologue  in  A.  abditus,  but  whether  oral  papillae  of  all 
amphiurids  are  homologous  remains  to  be  seen.  The  ontogeny  of  A.  abditus  sup- 
ports A.  M.  Clark's  (1970)  contention  that  Amphioplus,  bearing  five  oral  papillae, 
occupies  a  central  position  among  the  amphiurids  such  that  the  oral  formulae  of 
other  genera  are  derived  by  a  simplification  of  the  Amphioplus  oral  armature. 
Further  studies  are  needed,  however,  to  decide  whether  this  complex  oral  structure 
is  a  primitive  or  an  advanced  trait. 


I  am  grateful  to  G.  V.  Irvine,  H.  R.  Lasker,  D.  L.  Meyer,  D.  L.  Pawson,  D. 
Schneider,  and  L.  P.  Thomas  for  editorial  criticism  of  various  drafts,  and  to  M. 
Downey  for  a  very  helpful  discussion.  I  am  indebted  to  H.  Sanders,  F.  Grassle, 
and  A.  Michael  for  providing  specimens  from  their  Wild  Harbor  collections.  This 
research  was  carried  out  at  the  University  of  Connecticut,  Woods  Hole  Oceano- 
graphic  Institution,  and  the  Smithsonian  Tropical  Research  Institute  (Galeta 
Marine  Laboratory).  I  thank  the  Librarians  at  those  institutions  and  the  Marine 
Biological  Laboratory  (Woods  Hole)  for  their  assistance,  and  also  acknowledge 
the  help  of  M.  B.  Abbott,  S.  Y.  Feng,  D.  R.  Franz,  R.  Gaskill.  and  C.  C.  Woo. 
This  research  was  supported  with  funds  from  an  NDEA  Title  IV  Fellowship,  a  Uni- 
versity of  Connecticut  Summer  Fellowship,  a  Woods  Hole  Oceanographic  Institu- 
tion Postdoctoral  Fellowship,  and  a  Walter  Rathbone  Bacon  Fellowship  (Smith- 
sonian Institution). 

SUMMARY 

1.  Amphioplus  abditus  has  a  vestigial  two-piece  larval  skeleton  that  has  portions 
with  different  crystallographic  orientations.     The  larval  skeleton  is  resorbed  and, 
unlike  that  of  echinoids,  it  does  not  act  as  a  center  of  formation  of  the  plates  of  the 
adult.     The  major  skeletal  elements  of  the   adult   develop  from   single    (usually 
triradiate)  spicules,  and  there  is  a  uniform  crystallographic  orientation  within  each 
plate. 

2.  The  radial  shields,  adoral  shields,  genital  plates  and  genital  scales  are  ophiu- 
roid  specializations  without  homologues  in  the  asteroids.    Ophiuroids  can  regenerate 
radial  shields  but  not  the  apical  primary  plates   (the  latter  are  probably  atavistic 
structures). 


OPHIUROID  SKELETAL  DEVELOPMENT  93 

3.  The  madreporite  and   oral   plates,   generally   thought   to  migrate   from   the 
dorsal  surface  of  the  disc,  originate  in  sit  it  on  the  ventral  surface  of  A.  abditus.     A 
dorsolateral  plate,  probably  confused  with  the  madreporite  in  past  studies,  is  a  pre- 
cociously formed  inter  radial- 1.     The  formation  of  a  "precocious  interradial  plate" 
could  be  a  vestige  of  the  primitive  ophiuroid  madreporite.    In  fact,  the  madreporites 
of  asteroids,  ancient  ophiuroids,  and  recent  ophuroids  may  not  be  homologous. 

4.  The  origin  of  each  of  the  oral  papillae  is  described.    Buccal  scales,  previously 
(and  incorrectly)  thought  to  develop  into  peristomial  plates,  form  the  second  oral 
papillae   in  A.    abditus.      Consequently,   the    second   oral    papillae    of   amphiurids 
should  not  be  considered  "oral  tentacle  scales".    The  true  tentacle  scales  are  cryptic 
structures  within  the  buccal  cavity. 

5.  The  oral  papillae  of  the  different  amphiurid  genera  are  probably  homologous. 
Judging  from  differences  in  the  oral  frame,  there  are  probably  two  major  amphiurid 
groups  composed  of  taxa  which  retain  the  buccal  scales  as  oral  papillae  (Amphio- 
plus and  possibly  Amphinra  ),  and  those  like  A.viognathus   (and  possibly  Awphi- 
pholis  and  Amphiodia)  which  resorb  the  buccal  scales. 

6.  A  new  system  of  homologues  is  suggested  for  the  plates  of  the  ophiuroid 
oral  skeleton.    The  proximal  oral  plate  is  considered  the  ambulacral  portion  of  the 
first  modified  arm-segment  and  buccal  scales  may  be  the  first  pair  of  adambulacrals. 
The  distal  oral  plates   (ambulacral),  adoral  shields   (adambulacral),  and  the  first 
ventral  arm-plate   (within  the  buccal  slit)   compose  the  second  transformed  arm- 
segment  of  the  oral  frame.    This  pattern  of  homology,  together  with  the  dissimilari- 
ties between  ophiuroid  and  asteroid  discs  constitute  important  differences  between 
the  ophiuroids  and  asteroids. 

LITERATURE  CITED 

APOSTOLIDES,  N.  C.,  1882.     Anatomic  et  developpement  des  Ophiures.     Arch.  Zool.  E.vp.  Gen., 

10:  121-224. 
BATHER,  F.  A.,  J.  W.  GREGORY,  AND  E.  S.  GOODRICH,  1900.     The  Echinoderma.  Pages  1-81   in 

E.  R.  Lankester,  Ed.,  A  treatise  on  zoology,  Part  3.   Adam  and  Charles  Black,  London. 
CLARK,  A.  M.,  1970.    Notes  on  the  family  Amphiuridae  (Ophiuroidea).     Bull.  Brit.  Mus.   (Nat. 

Hist.)  Zoo!.,  19:  1-81. 

CLARK,  H.  L.,  1914.     Growth-changes  in  brittle-stars.     Carnegie  lust.  Wash.  Publ..  182:  91-126. 
CUEXOT,   L.,    1948.     Anatomie,   ethologie   et   systematique   des   fichinodermes.      Pages    1-363   in 

P.  P.  Grasse,  Ed.,   Traite  dc  Zoologie  XL     Masson  et  Cie,  Paris. 
FELL,  H.  B.,  1941.     The  direct  development  of  a  New  Zealand  Ophiuroid..     Q.  J.  Microsc.  Sci., 

82:  377-441. 
FELL,    H.    B.,    1946.      Embryology   of   the   viviparous    ophiuroid   Amphipholis   squainata    Delle 

Chiaje.   Trans.  R.  Soc.  N.  Z.,  75  :  419-464. 
FELL,  H.  B.,   1963.     The  phylogeny  of  sea-stars.     Phil.   Trans.  R.  Soc.  London  Scr.  B.,  246: 

381-436. 
FEWKES,  J.  W.,  1887.     On  the  development  of  the  calcareous  plates  of  Amphiura.     Bull.  Mus. 

Comp.  Zool.  Harvard  Univ.,  13:  107-150. 

GRAVE,  C.,  1900.     Ophiura  brcrispina.     Mem.  Biol.  Lab.  Johns  Hopkins  Univ..  4:   79-100. 
GUILLE,  A.,  1964.     Contribution  a  1'etude  de  la  systematique  et  de  1'ecologie  d'Ophiothri.r  quin- 

qiteinaculata  d.  Ch.  Vie  Milieu,  15:  243-301. 
HENDLER,  G.,  1973.     Northwest  Atlantic  amphiurid  brittlestars,  Amphioplus  abditus   (Verrill), 

Amphioplus    macilcntus    (Verrill),    and    Amphioplus    scpultus    n.    sp.     (Ophiuroidea: 

Echinodermata)  :     systematics,   zoogeography,   annual   periodicities,   and   larval   adapta- 
tions.     Ph.D.    Dissertation,    University    of    Connecticut,    Storrs,    Connecticut,    255    pp. 

(Diss.  Abstr.,  34:  3106-B;  order  no.  74-00007.) 


94  GORDON  HKNTJLER 

I  I  I:\DLEU,   G.,   1975.     Adaptational   significance'  (if  tin-  patterns  of  dphiuroid  dcveldpnicnt.   Am. 

Zoo!.,  15:  691-715. 
HENDLER,  G.,  1977.    Development  of  Amphioplus  abditus  ( Yen-ill)    (Echinodermata :  Ophiuroi- 

dea)  :  I.  Larval  biology.    B'wl.  Bull.,  152:  51-  63. 
HUMASON,    G.    L.,    1967.      Animal    tissue    techniques.     W.    H.    Freeman    and    Company,    San 

Francisco,  569  pp. 
HYMAN,  L.  H.,  1955.     The  invertebrates:    Echinodermata.    McGraw-Hill  Book  Company,  New 

York,  763  pp. 
KROHN,  A.,  1851.     Ueber  die  Entwickelung  einer  lebendig  gebarenden  Ophiure.     Arch.  Aunt. 

PhysioL,  1851:  338-343. 

LUDWIG,  H.,  1878.    Beitrage  zur  Anatomic  der  Ophiuren.   Z.  JViss.  ZooL.  31  :  346-394. 
LUDWIG,  H.,   1881.     Zur  Entwicklungsgeschichte  des  Ophiurenskelettes.     Z.    Jl'iss.  ZooL,  36: 

181-200. 

LUDWIG,  H.,  1899.    Jugendformcn  von  Ophiuren.    Sber.  Prcitss.  Akad.  U'iss.,  14:  210-235. 
LUDWIG,  H.,  1901.     Echinodermen  III.     Die  Schlangensterne.     Pages  745-966  in  H.  G.  Bronn, 

Ed.,  Klassen  and   Ordnungcn   des   Thicr-Rcichs,   Band  2,   Abt.   3,    C.    F.   Winter'sche 

Verlagshandlung,  Leipzig. 
MACBRIDE,  E.  W.,  1906.     Echinodermata.     Pages  425-623  in  S.  F.  Harmer  and  A.  E.  Shipley, 

Eds.,  The  Cambridge  Natural  Historv,  Vol.  I,  MacMillan  and  Co.,  Ltd.,  London. 
MORTENSEN,  T.,  1912.    UbeV  Astcronyx  loveni  M.  Tr.  Z.  U'iss.  Zoo!.,  101:    264-289. 
MORTENSEN,  T.,  1913.     Die  Echinodermenlarven  der  Deutsche  Siidpolar-Expedition  1901-1903. 

Dent.  Siidpol.-E.rpcd.,  14:  69-111. 
MORTENSEN,  T.,  1921.     Studies  on  the  development  and  larval  forms  of  echinodcrms.    G.  E.  C. 

Gad,  Copenhagen,  261  pp. 
MORTENSEN,   T.,    1931.     Contributions   to   the   study   of   the   development   and   larval    forms   of 

echinoderms.     I.  The  development  and  larval  forms  of  some  tropical  echinoderms.    II. 

Observations   on   some   Scandinavian   echinoderm   larvae.     Kgl.   Dan.    Vidensk.   Sclsk. 

Skr.  Ser.  9,  4  :  1-39. 
MORTENSEN,  T.,   1933a.     The  echinoderms  of   St.   Helena.     Vidensk.  Mcdd.  Naturliist.  Foren. 

Kjobcnhavn,  93 :  26-73. 

MORTENSEN,  T.,  1933b.    Ophiuroidea.    Dan.  Itu/olf-E.rpcd.,  8:  1-121. 
MORTENSEN,  T.,  1936.    Echinoidea  and  Ophiuroidea.    Discovery  Rep.,  12:  199-348. 
MORTENSEN,   T.,    1938.     Contributions   to   the   study   of   the   development   and   larval    forms   of 

echinoderms  IV".    Kgl.  Dan.  Vidensk.  Sclsk.  Skr.  Naturv.  Math.  Scr.  9,  7  :  1-59. 
MULLER,  J.,   1851.     Ueber  die  Ophiurenlarvaen  des  Adriatischen  Meres.    Abh.  Preuss.  Akad. 

Wiss.,  1851:  33-61. 
MURAKAMI,    S.,    1937.     On   the  development   of  the   calcareous   plates   of   the   ophiuroid   larva, 

Ophiopluteus  scrratus.    Annot.  ZooL  Jpn.,  16:  135-147. 
MURAKAMI,  S.,  1940.    On  the  development  of  the  calcareous  plates  of  an  ophiuran,  Amphipholis 

japonica  Matsumato.   Jpn.  J.  ZooL,  9:  19-33. 
MURAKAMI,  S.,  1941.     On  the  development  of  the  hard  parts  of  a  viviparous  ophiuran,  Stego- 

phinra  sculpta  (Duncan).     Annot.  ZooL  Jpn.,  20:  67-78. 
OLSEN,  H.,   1942.     The  development  of  the  brittle-star   Ophiopholis  aculeata   (O.   Fr.   Miiller), 

with  a  short  report  on  the  outer  hyaline  layer.     Bcrgcns  Mus.  Arb.  Natun-id.,  1942: 

1-107. 
ONODA,  K.,  1931.     Notes  on  the  development  of  Hcliocidaris  crassispina  with  special  reference 

to  the  structure  of  the  larval  body.    Kyoto  Imp.  Univ.  Mem.  Coll.  Sci.  B,  7:  103-134. 
SCHOENER,  A.,  1967.     Post-larval  development  of  five  deep-sea  Ophiuroids.     Deep  Sea  Res.,  14 : 

645-660. 
SCHUCHERT,    C.,    1915.      Revision   of   paleozoic    Stelleroidea   with    special    reference    to    North 

American  Asteroidea.    Bull.  U.  S.  Nat.  Mus.,  88:  1-311. 
SIMROTH,  H.,  1876.    Anatomic  und  Schizogonie  der  Ophiactis  virens  Sars.  Z.  U'iss.  ZooL,  27: 

419-485. 
SOLLAS,  I.  B.  J.,  AND  W.  J.  SOLLAS,  1912.     Lapworthura:  a  typical  brittlestar  of  the  Silurian 

Age;  with  suggestions  for  a  new  classification  of  the  Ophiuroidea.   Phil.  Trans.  R.  Soc. 

London  Ser.  B,  202  :  213-232. 
SPENCER,  W.   K.,   1951.  Early  paleozoic  starfish.     Phil.    Trans.  R.  Soc.   London  Ser.   B,  235: 

87-128. 


OPHIUROID  SKELETAL  DEVELOPMENT  ^5 

SPENCER,  \V.  K.,  AND  C.  \V.  WRIGHT,  1966.  Asterozoans.  Pages  5-107  in  R.  C.  Moore,  Ed., 
Treatise  on  invertebrate  palet>ntt>l«</y.  I'urt  I'.  Echinodermata  3.  I'ul.  1.  The  Uni- 
versity of  Kansas  Press,  Lawrence. 

THOMAS,  L.  P.,  1962.  The  shallow-water  amphiurid  brittle  stars  (Echinodermata,  Ophiuroidea) 
of  Florida.  Bull.  Mar.  Sci.,  12:  623-694. 

TURNER,  R.  L.,  AND  J.  H.  DEARBORX,  1972.  Skeletal  morphology  of  the  mud  star,  Ctcnodiscus 
crispatus  ( Echinodermata  :  Asteroidea  ) .  /.  Morphol.,  138  :  239-262. 

ZUR  STRASSEN,  O.,  1901.  Zur  Morphologic  des  Mundskelettes  der  Ophiuriden.  /.ool.  ./».;.,  24: 
609-620. 


Reference:   Hwl.  Hull..  154:  96-120.      (February, 


A  GENERIC:  REVISION  OF  TIIK  BRACKISH-WATER  SERPULID 

FICOPOM.l  Tl AS  SOUTHERN  1921  (POLYCHAETA  :  SERPULINAE  ), 

INCLUDING  MERCIERELLA  FAUVEL  1923,  SPHAEROPOMATUS 

TREADWELL  1934.  MERCIERELLOPSIS  RIO]  A  1945  AND 

NEOPOMATUS  PILLAI  1960 

H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

Laboratory  for  Zoological  Ecology  and  Taxonomy,  State  University  of  Utrecht, 
Plomfctorengracht  9-11,  Utrecht,  The  Netherlands 

In  the  last  half  century,  five  monotypic  serpulid  genera  have  been  described 
exclusively  from  brackish  waters,  and  Pillai  (  1960)  has  united  four  of  them  in  the 
subfamily  Ficopomatinae  (see  below).  One  of  these,  Mcrcierclla,  has  received  the 
attention  of  many  biologists  in  various  fields  of  research.  There  has  been  consider- 
able confusion  about  the  identity  of  two  of  the  species,  namely,  Mcrcierella  enigma- 
tica  and  Neoponiatiis  nschakori  (see  lists  of  synonyms  in  this  paper).  This 
confusion,  and  the  view  of  the  senior  author  that  it  was  unlikely  that  similar  special- 
izations in  the  brackish  habitat  were  evolved  by  five  different  but  evidently  closely 
related  genera,  made  a  review  necessary.  Preliminary  work  was  done  by  the  senior 
author  and  later  elaborated  by  the  junior  author  as  partial  fulfillment  of  his  post- 
graduate studies. 

MATERIALS  AND  METHODS 

The  greater  part  of  the  material  came  from  the  collections  of  the  British  Museum 
(Natural  History)  and  of  the  senior  author,  from  whose  material  small  series  have 
been  presented  to  other  museums.  Several  institutions  sent  material  as  gifts  or  loans. 

The  photographs  were  taken  by  Mr.  E.  van  der  Vlist  and  other  staff  at  the 
Zoologisch  Laboratorium,  Rijksuniversiteit,  Utrecht.  Drawings  of  opercula  were 
made  by  using  a  drawing-prism.  In  order  to  make  camera-lucida  drawings,  it  was 
necessary  to  separate  the  setae  and  uncini.  This  was  achieved  by  putting  entire 
animals  in  a  few  drops  of  10%  KOH  for  12-24  hours  and  subsequently  squashing 
them  in  glycerin-gelatin.  Some  figures  were  drawn  from  tufts  of  setae,  extirpated 
from  the  animals,  and  preserved  in  glycerin-gelatin.  All  figures  were  drawn  by  the 
junior  author. 

Measurements,  unless  stated  otherwise,  are  given  in  mm  ;  meristic  values  are 
based  upon  counts  in  ten  specimens  minimally,  unless  otherwise  stated  (e.g.,  n  = 
3).  References  in  synonymies  preceeded  by  " [  ?] "  indicate  a  questionable  identifica- 
tion in  our  view.  The  following  abbreviations  are  used  for  collections :  AHF, 
Allan  Hancock  Foundation,  University  of  Southern  California,  Los  Angeles ;  AM, 
the  Australian  Museum,  Sydney;  AMNH,  the  American  Museum  of  Natural 
History,  New  York;  BMNH,  British  Museum  (Natural  History),  London; 
MNHN,  Musee  Nationale  d'Histoire  Naturelle,  Paris;  RMNH,  Rijksmuseum  van 
Natuurlijke  Historic,  Leiden;  SME,  Station  Marine  d'Endoume,  Marseille;  tHU, 

96 


REVISION  OF  FICOPOMATUS 


97 


FIGURE  1.  Opercula,  different  orientations.  The  specimens  represented  in  a-e  are  Ficopo- 
iiiahis  macrodon  from  Taleh-Sap ;  the  series  shows  the  possible  development  from  young  to  old. 
The  specimens  in  f-i  are  F.  inuunicnsis:  /,  syntype  from  Miami;  g,  from  Curacao;  h-i,  from 
Barbados,  with  and  without  horny  plate.  All  are  to  the  same  scale. 


collection  of  H.  A.  ten  Hove,  Utrecht;  USNM,  National  Museum  of  Natural 
History,  Smithsonian  Institution  Washington  (formerly  United  States  National 
Museum)  ;  ZMA,  Instituut  voor  Taxonomische  Zoologie,  Zoologisch  Museum, 
Amsterdam ;  and  ZMU,  Laboratorium  voor  Zoologische  Oecologie  en  Taxonomie, 
Zoologisch  Museum,  Utrecht. 


98  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

TAXONOMY 

Genus  Ficoponnittis  Southern,  1921 

Type-species:  Ficopomatus  nutcrodon  Southern,  1921,  by  monotypy.  Gender: 
masculine.  Synonyms:  Mercierclla  Fauvel,  1923,  type-species:  M.  enigmatica 
Fauvel,  1923,  by  monotypy;  Sphaeropomatus  Treadwell,  1934,  type-species:  S. 
iniainicHsis  Treadwell,  1934.  by  monotypy;  Mer tier ellop sis  Rioja,  1945,  type- 
species:  M.  prietoi  Rioja,  1945.  by  monotypy;  Ncopoinahts  Filial,  1960,  type- 
species:  N.  itscliokori  Pillai,  1960,  by  original  designation. 

Original  diagnosis:  "Modified  setae  present  on  the  first  thoracic  segment,  hav- 
ing blades  provided  with  very  stout  teeth.  Beneath  the  blades  is  a  transverse  row 
of  more  than  two  teeth.  Uncini  with  relatively  few  teeth,  the  lowest  of  which  is  in 
the  form  of  an  elongate  bifid  spine.  Ventral  abdominal  setae  geniculate.  Operculum 
fig-shaped,  without  any  outgrowths"  (Southern,  1921,  p.  655). 

Emended  diagnosis :  Tube  white,  gradually  increasing  in  diameter  toward 
anterior  end  and  semicircular  in  cross-section.  One  or  three  keels  sometimes  present. 

Thoracic  segments  seven,  with  six  uncinigerous.  Collar  setae  coarsely  serrated 
and  limbate.  Remaining  thoracic  setae  limbate.  Thoracic  uncini  saw-like,  excep- 
tionally partly  rasp-like,  with  six  to  twelve  teeth  visible  in  profile,  including  anterior 
gouged  tooth.  Uncinigerous  tori  placed  in  two,  nearly  parallel  rows.  Abdominal 
setae  geniculate  with  denticulate  edge.  Abdominal  uncini  saw-  or  rasp-like,  with 
one  to  four  rows  of  teeth,  six  to  fourteen  teeth  visible  in  profile,  including  anterior 
gouged  tooth.  Posterior  abdominal  segments  without  capillary  setae  and  without 
dorsal  glandular  area. 

Operculum  consisting  of  bulbous  fleshly  part,  terminated  by  horny  plate ;  pe- 
duncle smooth,  without  filaments  or  wings,  inserted  just  below  left  branchial  lobe, 
near  medial  line.  No  pseudo-operculum  present. 

Collar  not  lobed,  with  entire  edge,  continuous  with  thoracic  membranes  which 
are  united  ventrally  on  anterior  abdominal  segments.  Branchial  filaments  arranged 
in  two  semicircles,  not  united  by  branchial  membrane.  Pair  of  ventral  mouth- 
palps  absent. 

Key  to  species  of  Ficopomatus 
1.         Operculum  not  "spiny"  ( Fig.  1)    2 

Operculum  "spiny"   (Fig.  2)    3 

2(1).  Operculum  with  conical  horny  cap  with  dorsal  furrow   (Fig.  la-e)  ;  tube  usually  \vith 

median  keel   ( Fig.  5e)    F.  inacrodon 

Operculum   without   horny   endplate,   or   with   slightly   concave   one    (Fig.    If-i)  ;   tube 

without  median  keel   ( Fig.  5a,  b )    F.  iniamicnsis 

3(1).  Operculum  distally  convex,  "spines"  curved  outwards   (Fig.  2a-d)  ;  thoracic  membranes 

fused  dorsally    F.   uschakovi 

Operculum  distally  concave,  "spines"  curved  inwards   (Fig.  2f-i )  ;  thoracic  membranes 
not  fused  dorsally   F.  cnigmaticus 

Discussion:  As  already  stated  by  Southern  (1921,  p.  655)  the  main  character 
separating  Ficopotnatns  from  all  other  known  serpulid  genera  is  the  peculiar  shape 
of  the  special  collar  setae.  A  presumed  difference  in  collar  setae  and  presence  of 


REVISION  Ol'   r-'ICOl'OMATUS 


FIGURE  2.  Opercula,  different  orientations.  The  specimens  represented  in  a-d  are  Flcopo- 
matus  uschakovi:  a-c,  from  Guadalcanal;  d,  paratype  of  var.  lingayancnsis  from  Luzon.  The 
specimens  in  e-i  are  F.  enigmaticus:  e-h,  from  the  Netherlands;  i,  from  Uruguay.  Scale  in  f, 
g,  h  is  2  mm;  in  remaining  figures,  the  scale  is  0.5  mm. 

spines  on  the  operculum  were  the  main  reasons  for  Fauvel's  (1923,  p.  429)  propos- 
ing the  new  genus  Mercierella  for  his  specimens.  Erroneously,  Treadwell  ( 1934,  p. 
340)  counted  only  six  thoracic  setigers,  which,  along  with  the  presumed  difference 
in  collar  setae,  was  his  main  reason  for  erecting  Sphaeropomatiis  .  Rioja  (1945,  pp. 
412-413)  acknowledged  the  similarity  of  his  Mercierellopsis  with  Sphaeropomatus, 
but  thought  it  to  he  different  in  the  number  of  thoracic  setigers  and  in  the  presence 
of  a  horny  endplate  on  the  operculum  of  the  former  ( not  mentioned  for  Sphaero- 
pomatus by  Treadwell,  although  found  to  be  present  on  his  material).  Similar 
reasons  finally  led  Pillai  (1960,  p.  33)  to  describe  a  fifth  genus,  Xcopomatits, 
although  he  recognized'  the  similarities  among  four  of  the  five  genera  by  creating 
the  subfamily  Ficopomatinae  for  them  (1960,  p.  35).  All  differential  diagnoses, 


100 


H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 


FIGURE  3.  Branchial  filaments  (a-e)  ;  thorax  (f)  ;  and  variability  in  upercular  spines 
( g-q ) .  The  specimens  represented  are :  a,  f-k,  Ficopomatus  uschakovi  from  Guadalcanal ;  b,  F. 
macrodon  from  Taleh-Sap;  c,  F.  iniainicnsis  from  Jamaica;  and  d,  e,  1-q,  F.  cnigmaticus  from 
the  Netherlands. 

given  by  these  authors,  were  based  upon  literature  only. 

Straughan  (1966,  p.  145)  held  the  opinion  that  ".  .  .  there  is  a  continuous  cline 
between  isolated  populations  between  Sydney,  where  the  brackish  water  serpulids 
are  typical  of  Mercierella,  and  Brisbane,  where  the  brackish  water  serpulids  are 
typical  of  Ncopomatus  .  .  ."  For  that  reason  she  synonymized  Neopomatns  uscha- 
kovi with  Mercierella  enigmatica.  This  has  been  refuted  by  Hartmarm-Schroder 
(1971)  and  Pillai  ( 1971  ).  Although  both  authors  report  an  enormous  infraspecific 
variability,  they  still  maintain  a  generic  separation  of  the  two  species.  Pillai  (1971) 
redescribed  the  genera,  with  the  omission  of  Mercierellopsis. 

Thus,  most  authors  agree  that  there  are  many  similarities  in  this  group,  but 
that  there  are  five  different  genera  based  on  differences  in  denticulation  of  collar 
setae,  the  presence  of  spiny  or  nonspiny  opercula,  differences  in  shape  of  collar  and 
thoracic  membranes,  and  presence  or  absence  of  "peristomes"  on  the  tubes  (collar- 
like  rings  which  indicate  the  position  of  former  peristomes). 


REVISION   OF  FICOPOMATUS  101 

With  regard  to  the  differences  in  collar  setae,  a  considerable  variation  in  their 
shape  may  occur  within  one  species,  even  in  one  specimen  (e.g..  ten  Hove,  1974, 
Figs.  4—9,  for  Hydroidcs  norvegica  Gunnerus).  Extreme  variability  in  opercula 
was  noted  by  ten  Hove  (1970,  Figs.  77,  121,  123)  for  Spirobranchus  polyccnis 
(Schmarda)  and  (1975,  plates  I-III)  tor  Pseudovermilia  occidcntalis  (Mclntosh). 
Collar  and  thoracic  membranes  are  thin,  fleshy  structures  ;  their  shape  is  dependent 
on  the  method  of  preservation.  The  presence  or  absence  of  "peristomes"  may  de- 
pend upon  environmental  conditions  (Hartmann-Schroder,  1967,  p.  454,  for  Mer- 
cierclla  enit/inatica )  ;  a  further  example  of  variability  of  tubes,  to  the  extent  that 
within  one  species  "peristomes"  may  be  present  or  absent,  is  given  by  ten  Hove 
(1973,  plates  I-II)  for  three  species  of  Sclcrostyla  and  (1975,  plate  VII)  for 
Pseudovermilia  occidentalis. 

Considering  that  the  differences  between  the  five  "genera"  under  discussion  are 
certainly  not  greater  than  the  above-mentioned  examples  of  variability,  and  con- 
sidering the  striking  similarities  summarized  in  our  emended  generic  diagnosis,  it 
does  not  seem  realistic  to  consider  them  as  distinct.  Further  arguments  for 
synonymizing  these  genera  will  be  given  in  the  various  discussions  following  the 
species  descriptions.  In  our  opinion,  the  differences  between  Ficopomatus  and 
other  serpulid  genera  are  too  small  to  justify  a  distinction  on  the  subfamily  level. 
Therefore,  we  suggest  that  the  subfamily  Ficopomatinae  Pillai  (1960,  p.  35)  be 
withdrawn  from  recognition. 

All  species  of  the  genus  Ficopomatus  may  occur  as  solitary  individuals  or  in 
dense  aggregated  masses.  A  discussion  of  the  possible  causes  of  mass  occurrence 
has  been  given  by  ten  Hove  (1978). 

Ficopomatus  capensis  Day  (1961,  pp.  552-553,  Fig.  17  h-n ;  1967,  pp.  810-812, 
Fig.  38.5  j-n)  definitely  cannot  be  included  in  the  genus,  as  emended  above.  From 
the  figures  and  description  it  more  probably  should  be  placed  in  Chitinopoma 
Levinsen,  emended  (Zibrowius,  1969),  Chitinopomoides  Benham,  or  Pseudo- 
chitinopoma  Zibrowius.  Since  these  three  genera  are  mainly  characterized  by  the 
microstructure  of  the  setae  and  uncini,  the  generic  position  of  Day's  species  can 
only  be  established  after  a  careful  comparison  of  material  of  the  genera  concerned. 

The  occurrence  of  wide,  flaring  "peristomes"  on  the  tubes  of  fossil  serpulids  was 
thought  to  be  of  generic  diagnostic  value  by  some  palaeontologists.  This  has  been 
disputed  by  Hartmann-Schroder  (1967,  p.  452).  To  our  knowledge,  '"peristomes" 
may  occur  in  species  of  the  genera  Chitinopomoides  Benham.  Cntcigera  Benedict. 
Filograna  Oken,  Josepliclla  Caullery  and  Mesnil,  Metavennilia  Bush.  Pseitdorcr- 
inilia  Bush.  Scrpiila  Linnaeus,  and  rcnniliopsis  Saint-Joseph.  The  tubes  of  Serpula 
narconensis  Baird,  as  figured  by  Mclntosh  (1885.  plate  54,  Fig.  5)  are  very  similar 
to  those  of  Ficopomatus  enigmaticus.  Judging  by  the  figures  and  measurements, 
Mercierella?  dubiosa  Schmidt  (1951,  p.  80,  Fig.  4)  might  belong  to  Filograna. 
For  similar  reasons,  the  tube  of  M.  rorcretoi  Schmidt  (1951,  p.  80,  Fig.  5)  is 
strongly  reminiscent  of  a  yet  undescribed  species  of  Serpula  from  the  Caribbean. 
Regenhardt  (1961)  erected  a  genus,  Proliserpula,  containing  species  with  "peri- 
stomes" and  suggested  a  connection  with  Mercierella.  In  our  opinion,  most  of  his 
species  resemble  recent  species  of  Pseudovermilia,  as  well  as  Filograna,  Serpula  and 
Vermiliopsis.  Mercierella  (?)  dacica  Dragastan  (1966,  pp.  147-150.  Figs.  1-3) 
resembles  not  only  Joscpliella.  but  also  some  calcareous  algae. 


102 


H.  A.  TEN  HOVE  AXD    T.   C.  A.   \VKKRDK\BURG 


XX 


nn  oo  pp 


yy 


zz 


qq      rr     ss 


tt    uu    vv 


FIGURE  4.    Collar  setae  are  shown  in  the  series,  a-n,  displaying  all  kinds  of  intergradations 
between  a  single  row  of  teeth  and  a  partial  triple  row ;  however,  a  prominent  smooth  gap  be- 


REVISION  OF  FICOPOMATUS  103 

Ficopomatus  macrodon  Southern  1921 
Figures  la-e ;  3b ;  4e-g,  o-p,  t-u,  cc-dd,  ww ;  5e 

Serpulid — Annandale  1916,  p.  100  [Siam,  Taleh-Sap;  material  studied  by  us]. 
Ficopomatus   macrodon    Southern,    1921,    pp.    655-659,    plate    30,    Fig.    27   A-M 

|  India,  Cochin  Backwater;  extensive  description];  Rioja,  1924,  pp.  168-169 
[no  new  data;  comparison  with  F.  enigmaticus];  Mclntosh,  1926,  pp.  405-423, 
plate  13,  Fig.  3,  plate  15,  Fig.  3  [India,  Chilka  Lake;  no  new  data;  anatomical 
description  of  operculum]  ;  Hartman,  1959,  p.  575  [name  only]  ;  Pillai,  1971,  pp. 
115-116,  Fig.  8A-F  [Ceylon,  Tambalagam  Lake;  description  and  comparative 
study]. 

[?]Ficopomatus  macrodon — Fauvel,  1931,  p.  1069  [India,  Madras  Coast,  Ennui- 
Backwater;  name  only,  could  be  P.  •nschakot'i;  see  discussion  below]  ;  Fauvel, 
1932,  pp.  248-249  [India,  Madras  Coast,  Ennur  Backwater;  Sunderbans;  Taleh- 
Sap,  Gulf  of  Siam,  Stats.  11,  17,  21,  29,  and  32;  description;  some  material  of  Stat. 
32  studied  by  us;  see  discussion  below]  ;  Fauvel,  1953,  pp.  473-474,  Fig.  248  c-1 
[India,  Madras  Coast,  Ennur  Backwater;  Cochin  Backwater;  Chepparan;  Sunder- 
bans; Taleh-Sap,  Gulf  of  Siam;  description;  see  discussion  below]. 
Ficopomatus — Pillai,  1960,  pp.  32-35  [comparative  study;  diagnosis]. 

nonFicopoinatiis  sp. — Hill,  1967,  pp.  303-321  [Nigeria,  Lagos;  name  only;  most 

likely  abnormal  /'".  uschakovi;  see  discussion  below]. 
Mcrcierclla  enigmatica — Nelson-Smith,  1967,  p.  54  (in  part?)    [the  palaeotropical 

records  most  probably  are  of  F.  macrodon  or  F.  uschakovi;  the  diagnosis  and 
figures  are  of  F.  enigmaticus] . 
[l]Mercierella  enigmatica — Ganapati,  Lakshmana  Rao,  and  Nagabhushanam,  1958, 

pp.  197-206  [India,  17°N;  83°E;  name  only,  material  is  probably  F.  macrodon 
or  F.  uschakovi] . 

Material  Studied.  Thailand:  Taleh-Sap,  Annandale  Collection,  Stat.  32  (five 
isolated  specimens  and  three  small  pebbles  with  seven  specimens  in  tubes  [BMNH 
1938:  5:  7:  89-91];  ten  specimens  identified  by  P.  Fauvel  and  H.  Zibrowius 
[MNHN]). 

Tube :  The  tube  is  shining  white,  semicircular  in  cross-section.  Collar-like 
rings  of  former  peristomes  were  absent  in  the  material  studied.  Most  of  the  tubes 
have  a  high  and  sharp  median  keel  (Fig.  5e)  ;  however,  in  a  few  tubes  it  is  in- 
distinct. 

Branchiae :  The  branchial  filaments  arise  from  paired  lobes  and  number  about 
six  (5-7;  n  =  5)  on  the  left  and  seven  (6-7;  n  =-  5)  on  the  right.  They  are 
arranged  in  two  semicircles  and  are  not  connected  by  a  branchial  membrane.  The 
filaments  are  shorter  ventrally,  their  tips  being  free  of  pinnulae  to  a  greater  or 
lesser  extent  (Fig.  3b). 

tween  proximal  coarse  bunches  of  teeth  and  distal  series  of  teeth  is  nowhere  present.  Posterior 
setae  are  shown  as  follows :  o-s,  other  thoracic  setae ;  t-bb,  thoracic  uncini ;  cc-vv,  abdominal 
uncini  (cc-gg,  jj-kk,  nn-pp,  anterior  segments;  qq-ss,  middle  segments;  hh-ii,  11-mm,  tt-vv, 
posterior  segments);  and  ww-zz,  abdominal  setae.  Species  represented  are:  F.  enigmaticus 
from  the  Netherlands,  a-d,  s,  aa-bb,  nn-vv,  7.7. ;  F.  macrodon  from  Taleh-Sap,  e-g,  o-p,  t-u,  cc- 
dd,  wvv ;  F.  miamiensis  from  Barbados,  h-i,  q,  v-\v,  ee-ii,  xx;  and  F.  uschakoi'i  from  Java, 
j  ;  from  Guadalcanal,  k-n,  r,  x-z,  jj-mm,  yy.  All  are  to  the  same  scale. 


104 


H.  A.  TEN  HOVE  AND  J.  C  A.  WEERDENBURG 


10 


FIGURE  5.  Tubes  of  Ficopomatus:  a,  b,  F.  miamiensis  from  Barbados,  showing  differences 
between  two  populations  from  Holetown  river  pool  and  from  onr-half  mile  north  of  Bcllairs 
Institute ;  c,  F.  enigmaticus  from  the  Netherlands ;  d,  F.  uscliakovi  from  India,  showing  three 
longitudinal  keels ;  and  e,  F.  macrodon  from  Taleh-Sap,  showing  one  longitudinal  keel. 


REVISION  OF  FICOPOMATUS  105 

Peduncle:  The  peduncle  is  smooth,  sometimes  faintly  wrinkled,  especially  just 
below  the  bulb  of  the  operculum.  It  is  subtriangular  in  cross-section  with  a  shallow 
dorsal  groove.  There  is  a  gradual  transition  train  peduncle  to  opercular  bulb 
(Fig.  la-b). 

Operculum:  The  operculum  is  a  fleshly  bulb,  terminated  by  a  more  or  less 
conical  horny  cap,  which  has  a  dorsal  furrow  (Fig.  la-e).  The  thicknr^  of  the 
horny  plate  is  positively  correlated  with  increasing  length  of  the  cone. 

Collar  and  thoracic  membranes  :  The  collar  is  rather  high,  not  lobed,  and  has  an 
entire  edge.  It  is  continuous  with  the  thoracic  membranes,  which  are  united  ven- 
trally  on  the  anterior  abdominal  segments. 

Thorax  :  The  thorax  has  seven  segments,  six  of  which  are  uncinigerous.  The 
bundles  of  collar  setae  contain  only  a  few  setae  of  two  types  :  coarsely  serrated  ones 
(Fig.  4e-g;  see  discussion  below)  and  limbate  ones.  Subsequent  fascicles  of 
setae  are  larger  and  are  in  two  nearly  parallel  rows,  containing  limbate  setae  only 
(Fig.  4o-p).  The  thoracic  uncinigerous  tori  are  arranged  in  two  nearly  parallel 
rows,  with  about  70  ( n  ==2)  uncini  per  torus.  The  thoracic  uncini  have  a  single 
row  of  teeth  ;  however,  directly  above  the  most  anterior  gouged  tooth,  there  are 
one  or  two  transverse  rows  of  two  or  three  teeth  (Fig.  4t-u).  There  are  ten  to 
twelve  (  n  =  •  4  i  teeth  visible  in  profile.  Thoracic  uncini  from  the  first  row  do  not 
differ  essentially  from  those  in  the  last  row. 

Abdomen  :  Owing  to  the  scanty  and  incomplete  material,  the  number  of  ab- 
dominal segments  and  uncini  per  row  could  not  be  determined.  The  abdominal 
uncini  are  all  rasp-like,  with  three  to  four  rows  of  small  curved  teeth;  about  13-15 
teeth  (n  ==  3)  are  visible  in  profile,  including  the  anterior  gouged  one  (Fig.  4cc- 
dd).  The  bundles  of  abdominal  setae  consist  of  two  to  five  geniculate  ones  (Fig. 
4w  w ) . 

Size  :  The  length,  including  operculum,  is  at  least  up  to  7  mm,  but  cannot  lie- 
given  more  exactly  owing  to  the  incompleteness  of  the  specimens;  the  width  of  the 
thorax  is  usually  0.5-0.7  mm  (n=  4).  The  branchiae  and  the  operculum  may 
account  for  one-fifth  of  the  entire  length  of  the  animal. 

Discussion:  Unfortunately  the  type  material  of  F.  inacrodon  was  not  available, 
and  the  material  studied  was  in  poor  condition.  Yet  at  least  20  collar  setae  have 
been  studied. 

The  diagnostic  value  of  the  micro  structure  of  the  collar  setae  has  been  over- 
stressed  by  several  authors,  which,  along  with  some  other  variable  characteristics, 
has  resulted  in  five  different  genera.  According  to  Southern  (1921,  p.  656.  Fig. 
27D-E)  collar  setae  of  F.  inacrodon  have  a  transverse  row  of  teeth,  and  a  series  of 
very  coarse  teeth  distally,  separated  by  a  smooth  gap.  However,  Pillai  (1971,  p. 
116,  Fig.  8D-F )  indicated  that  the  above  mentioned  gap  occasionally  may  be 
absent.  In  our  material  this  gap,  if  present  at  all,  was  not  prominent.  Thus,  in  all 
probability  there  is  a  complete  cline  from  long  smooth  gap  to  continuous  series  of 
teeth.  Consequently,  this  cannot  be  used  as  a  distinguishing  characteristic.  More- 
over, Hartmann-Schroder  (1971,  Fig.  7c-d )  gives  a  similar  variation  in  the  collar 
setae  of  F.  iischakovi.  A  considerable  variation  in  shape  and  arrangement  of  teeth 
is  given  for  F.  cnit/tnaticits  by  Rioja  (1924,  Fig.  16-19)  and  Hartmann-Schroder 
(1967,  Fig.  4)  and  for  F.  uschakovi  by  Pillai  (1960,  Fig.  121-1).  I-K :  1965.  Fig. 
23G-I).  Our  studies  also  show  a  considerable  variation  in  the  microstructure  of 


106  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

the  collar  setae  (Fig.  4a-n)  and,  after  studying  more  than  one  hundred  slides,  each 
with  about  eight  special  setae,  we  still  are  incapable  of  distinguishing  the  species  by 
their  collar  setae  alone. 

F.  macrodon  can  be  distinguished  from  other  species  in  the  genus  by  its  peculiar 
thoracic  uncini.  In  our  opinion  this  feature  is  insufficient  to  justify  a  generic  dis- 
tinction (cf.,  Pseudovermilia  babylonia,  ten  Hove,  1975,  p.  96). 

The  operculum  of  F.  macrodon  resembles  those  of  some  specimens  of  F.  miami- 
ensis in  having  a  horny  endplate  without  spines.  However,  the  shape  of  this  end- 
plate  is  different  for  both  species  (cf.,  Fig.  1). 

Fauvel  (1931,  1932,  1953),  cited  by  Nelson-Smith  (1967,  p.  64),  mentions  both 
F.  macrodon  and  Mercierella  enigmatic®  from  Madras.  Fauvel  (1932,  p.  248)  states, 
"These  Ficopomatus  tubes  are  rather  square  in  section  with  three  dorsal  ridges, 
.  .  .  but  the  animals  enclosed  in  them  are  typical  Ficopomatus".  For  Mercierella 
enigmatica,  Fauvel  (1932,  p.  250)  reports,  "Its  tube  is  cylindrical,  ...  is  neither 
ridged  nor  enlarged  at  the  entrance."  However,  a  re-examination  of  part  of  his 
material,  from  this  locality,  labelled  Mercierella  enigmatica,  showed  tubes  with  three 
prominent  ridges,  containing  F.  uschakovi.  To  our  knowledge,  tubes  of  F.  macro- 
don generally  have  one  longitudinal  ridge  only.  It  is  evident  that  Fauvel's  descrip- 
tion is  confusing,  the  more  so  since  he  figures  European  material  (1953,  p.  475, 
Fig.  249),  and,  therefore,  his  identifications  need  to  be  checked. 

According  to  Annandale  (1916),  the  type-locality  Taleh-Sap  is  the  same  as  the 
inland  sea  of  Singgora  (also  spelled  Sengora  or  Songhkla).  It  is  located  on  the 
Malayan  peninsula,  on  the  Gulf  of  Siam  (see  Fig.  6).  It  appears  that  F.  macrodon 
occurs  in  brackish  waters  adjacent  to  the  Gulf  of  Bengal  and  the  Gulf  of  Siam. 

Ficopomatus  miamiensis  (Treadwell,  1934) 
Figures  If-i ;  3c ;  4h-i,  q,  v-w,  ee-ii,  xx ;  5a-b 

Sphaeropomatus  miamiensis  Treadwell,  1934,  pp.  338-341,  Figs.  1-5,  9  [Florida, 

Miami  River;  description;  syntypes  studied  by  us;  see  discussion  below]; 
Hartman,  1956,  p.  300  [Florida,  Indian  and  Miami  Rivers;  description;  material 
studied  by  us]  ;  Hartman,  1959,  p.  599  [name  only]  ;  Pillai,  1971,  pp.  116-119,  Figs. 
SG-H,  9  A-F  [studied  same  material  as  Hartman  (1956)  ;  extensive  description 
and  comparative  study;  see  discussion  below]  ;  Lacalli,  1977,  pp.  300-303,  Fig.  2 
[embryological  study  ;  material  studied  by  us] . 
Mer tier ellop sis  prietoi — Rioja,  1945,  pp.  411-417,  plates  1,  2  [Mexico,  Tecolutla 

(Gulf  of  Mexico);  extensive   description;   material  apparently   lost;   see   dis- 
cussion below];  Hartman,  1951,  p.   120   [no  new  data;  short  diagnosis;  see  dis- 
cussion below]  ;  Hartman,  1954,  p.  416  [name  only]  ;  Hartman,  1959,  p.  582  [name 
only ;  see  discussion  below] . 
Mercierella  enigmatica — Nelson-Smith,  1967,  p.  54  (in  part?)  [Curasao  is  listed  in 

the  distribution  of  F.  enigmaticus  ;  this  record  most  probably  should  be  referred 
to  F.  miamiensis;  the  diagnosis  and  figures  are  of  F.  enigmaticus  \. 

Material  studied.  United  States  of  America  :  Florida,  Miami  River,  17  May  1933, 
from  carapace  of  freshwater  shrimp,  Macrobrachiitm  jamaicensc  (Herbst),  Capt. 
John  W.  Mills  coll.  (18  specimens,  syntypes,  tubes;  USNM  20074,  20075,  20077; 
AMNH  2167;  tHU  210)  ;  Florida,  tributary  of  Indian  River,  Undersea  Institute  of 


REVISION  OF  PICOPOMATUS  107 

America  (six  specimens,  many  tubes;  AHF;  tHU  218)  ;  Florida,  Yero  Beach,  ad- 
jacent to  Indian  River,  artificial  ponds  at  Entomological  Research  Center,  22  and 
26  March  1963,   11  June  64   (40  specimens,  tubes;  USNM   54335-6);   Florida, 
Miami,  Coral  Gables  Canal,  25   March   1969,   M.  L.   Jones  coll.    (one  specimen; 
USNM   54337)  ;   Florida,   northwest  coast,   St.   Mark   Wildlife  Refuge,   near   St. 
Mark  Lighthouse  (30°  05'  N;  84°  12'  W),  15  Dec.  1976,  26  Feb  1977,  on  sub- 
merged tree  limbs  in  brackish  water  ponds,  salinity  \\%0,  P.  G.  Johnson  coll.  (28 
specimens,  tubes;  USNM  54795  ;  tHU  258)  ;  Louisiana,  Lake  Pontchartrain,  mouth 
of  industrial  canal,   salinity  2.5— 3%c,   M.A.   Poirrier  coll.    (eight  specimens,   tube 
fragments;  USNM   54794;  tHU  255).     Jamaica:    Great  Saltpond,  entrance  at 
Fort  Clarence,  8  May  1973,  P.  Wagenaar  Hummelinck  coll.,  Stat.  1681,  0-1  m 
depth  (50  specimens,  many  in  tubes;  BMNH  1976:  916-942;  tHU  234).     Bar- 
bados :  Holetown  River,  pool  near  bridge,  18  Feb.  1964,  P.  Wagenaar  Hummelinck 
coll.,  Stat.  1444   (four  specimens  and  pebbles  with  tubes;  RMNH   10706;  tHU 
235)  ;  about  1  km  North  of  Bellairs  Institute,  closed  lagoon,  encrusting  on  roots 
and  dead  branches,  April-May  1975,  T.  Lacalli  coll.  (six  specimens,  tube  fragments; 
tHU  225).    Curacao:   Bottom  of  H.N.L.M.S.  LUYMES,  after  one  month  in  Caribbean 
waters,  9  May  1970,  H.  A.  ten  Hove  coll.  (15  specimens  and  55  others  in  tubes; 
RMNH    10707;    SME)  ;    Schottegat,   east   of   Rijkseenheid    Boulevard,   opposite 
Zeelandia,  20  Sept.  1970,  11  Sept.  1975,  H.  A.  ten  Hove  coll.,  Stats.  2065,  2065a, 
limestone  boulders  in  sandy  mud,   Caulerpa,   10-20  cm  depth    (many  specimens, 
many  tubes;  tHU  254).    Belize  [==  British  Honduras]  :   Salt  Creek,  approximately 
8  km  north  of  Stann  Creek,  16  May  1977,  M.  L.  Jones  coll.,  in  channel  among 
mangroves,   about    1    m    depth,    on    living   Isognonwn   alatus,   temperature    32° C, 
salinity  3l%0  (four  specimens,  many  tubes;  USNM  54980;  tHU  259).    Panama: 
Canal  Zone,  Pacific  Third  Lock,  16  April  1972,  C.  E.  Dawson,  D.  L.  Pawson,  W. 
J.  Byas,  M.  L.  Jones  colls.,  USNM  Panama  Survey  Stat.  87-1,  cobbles,  rocks,  on 
shelf  adjacent  to  road   (24  specimens:  USNM  52743;  tHU  233);  Canal  Zone, 
Pacific  coast,  Upper  Miraflores  Lock  chamber,  26  Aug.  1974,  C.  E.  Dawson,  M.  L. 
Jones,  H.  \V.  Kaufman,  J.  Rosewater  colls.,  USNM  Panama  Survey  Stat.  203, 
lock  chamber  walls  (three  specimens;  USNM  54977-9). 

Tube :  The  tube  is  shining  white,  exceptionally  dull  and  roughened,  semicircu- 
lar in  cross-section.  There  are  no  longitudinal  ridges  or  keels.  Normally  collar- 
like  rings,  as  in  F.  enigmaticus,  are  absent  (Fig.  5a).  However,  in  the  popula- 
tions from  Barbados  (1  km  north  of  Bellairs  Institute),  Florida  (Vero  Beach), 
and  Panama  (Stat.  203),  wide  flaring  "peristomes"  are  present  (Fig.  5b). 

Branchiae :  The  branchial  filaments  arise  from  paired  lobes  and  number  about 
seven  (6-9)  on  the  left  and  eight  (6-10)  on  the  right.  They  are  arranged  in  two 
semicircles  and  are  not  connected  by  a  branchial  membrane.  The  filaments  are 
shorter  ventrally.  The  two  rows  of  pinnulae  become  shorter  txnvard  the  ends  of  the 
filaments,  which  are  free  of  pinnulae  to  a  greater  or  lesser  extent  (Fig.  3c). 

Peduncle :  The  peduncle  is  smooth,  sometimes  faintly  wrinkled,  and  circular  or 
subtriangular  in  cross-section.  There  is  a  gradual  transition  between  peduncle  and 
opercular  bulb  (Fig.  Ih). 

Operculum:  The  operculum  is  spherical  to  fig-shaped  (Fig.  lg-h),  sometimes 
with  a  horny  end-plate  (Fig.  If,  i),  which  may  be  flat  or  slightly  convex  (see  dis- 
cussion below).  The  operculum  never  has  spines. 


108  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

Collar  and  thoracic  membranes :  The  collar  is  high,  not  lobed  and  has  an  en- 
tire edge.  It  is  continuous  with  the  thoracic  membranes,  which  are  united  ven- 
trally  on  the  anterior  abdominal  segments. 

Thorax:  The  thorax  has  seven  segments,  six  of  which  are  uncinigerous.  The 
bundles  of  collar  setae  contain  only  a  few  setae  of  two  types :  coarsely  serrated  ones 
(Fig.  4h-i)  and  limbate  ones.  Subsequent  bundles  of  setae  are  larger  and  are  in 
two  nearly  parallel  rows,  containing  limbate  setae  only  (Fig.  4q).  Thoracic  un- 
cinigerous tori  are  arranged  in  two  nearly  parallel  rows  with  up  to  55  uncini  per 
torus.  The  uncini  along  the  entire  thorax  have  a  single  row  of  seven  (6-8)  curved 
teeth,  the  most  anterior  one  is  gouged  and  apparently  bifurcated  (Fig.  4v-w). 

Abdomen :  The  number  of  abdominal  segments  is  usually  about  40  (23-58, 
n  --  7).  The  anterior  two  or  three  segments  are  apparently  without  setae  or  un- 
cini. The  following  segments  have  very  few  uncini  (five  to  ten).  The  number  of 
uncini  per  row  slowly  increases  to  about  30  in  the  middle  of  the  abdomen,  then 
slowly  decreases  towards  the  pygidium  (about  three).  The  abdominal  uncini  of  the 
anterior  segments  are  partly  rasp-,  partly  saw-like,  in  such  a  way  that  within  a 
single  uncinus  both  conditions  may  occur  (Fig.  4ee-gg).  About  eight  to  ten  teeth 
are  visible  in  profile,  including  the  anterior  gouged  tooth.  The  uncini  of  the 
posterior  segments  are  smaller  and  rasp-like,  with  three  to  four  rows  of  small 
curved  teeth,  with  about  12  teeth  visible  in  profile,  including  the  anterior  gouged 
one  (Fig.  4hh-ii).  The  bundles  of  abdominal  setae  consist  of  three  (sometimes  one 
or  two)  geniculate  ones  (Fig.  4xx). 

Size:  The  length,  including  the  operculum,  is  about  7  mm  (2.5-11).  The 
width  of  the  thorax  is  about  0.8  mm.  The  branchiae  and  the  operculum  usually 
account  for  one-sixth  (sometimes  up  to  one-third)  of  the  entire  length  of  the 
animal. 

Discussion:  As  stated  above,  Treadwell's  (1934)  original  description  is  not 
entirely  correct;  his  main  errors  are  the  six  thoracic  setigers  (in  reality  seven)  and 
the  entirely  fleshly  operculum.  The  opercula  of  14  (out  of  18)  syntypes  did  show  a 
horny  endplate  (Fig.  If).  Of  about  200  specimens  studied  in  this  respect,  20% 
had  a  well-developed  endplate.  The  endplate  sometimes  is  difficult  to  see,  looking 
more  or  less  like  a  fleshy  brim.  Generally,  however,  the  endplate  is  missing  al- 
together. 

We  have  the  impression  that  there  is  no  relation  between  presence  or  absence  of 
endplate  and  the  size  of  the  specimens.  Pillai's  (1971,  Fig.  9A-C)  figures  are  based 
upon  collapsed  opercula  without  endplates  (material  re-examined).  Although 
Rioja  did  not  leave  a  collection  (according  to  a  personal  communication  from  Dr. 
Maria  Elena  Caso,  Institute  de  Biologia,  Universidad  Nacional  de  Mexico),  his 
figures  and  description  of  Mercierellopsis  prietoi  (1945,  pp.  411-417,  Figs.  1-20) 
are  excellent,  and  show  the  conspecificity  with  F.  niiainicnsis  beyond  doubt. 

In  contradistinction  to  all  previous  descriptions,  the  tube  may  show  wide  flaring 
"peristomes"  (Fig.  5b). 

Possibly  Morch's  (1863,  p.  353)  remark  on  the  occurrence  of  serpulid  tubes  on 
leaves  of  a  freshwater  plant  from  St.  Thomas  should  be  referred  to  F.  niiainiensis, 
although  some  spirorbids  can  occur  in  the  brackish  habitat  too. 

As  far  as  is  yet  known,  F.  miamiensis  is  restricted  to  Atlantic  tropical  and  sub- 
tropical areas  in  northern  and  middle  America,  and  a  more  or  less  isolated  locality 


REVISION  OF  FICOPOMATUS  109 

at  the  Pacific  end  of  the  Panama  Canal  (Fig.  6).  In  the  brackish  waters  of  Uru- 
guay and  Argentina,  it  is  replaced  by  F.  enigmaticus.  It  would  be  interesting  to 
to  know7  if  this  is  the  case,  too,  in  northern  America.  Since  the  only  record  of  F. 
enigmaticus  from  the  northern  Gulf  of  Mexico  is  from  the  bottom  of  a  boat,  it  is 
uncertain  if  this  represents  a  permanent  population. 

Ficopomatus  uschakovi  (Pillai,  1960) 
Figures  2a-d  ;  3a,  f-k  ;  4j-n,  r,  x-z,  j  j-mm,  yy ;  5d 

[  ?]  Serpuliden-Rohren  Ehlers,  1918,  p.  250   [Aru  Islands;  empty  tubes;  see  dis- 
cussion below]. 

Mercierella  enigmatica — Fauvel,  1931,  p.  1069  (in  part?)  [India,  several  localities; 
name  only;  the  record  of  Ennur  Backwater  is  F.  uschakoi'i  and,  perhaps,  F. 
macrodon,  as  well;  see  discussion  of  latter  species];  Fauvel,  1932,  pp.  249-251 
[India,  Madras  coast,  Ennur  Backwater;  description;  material  studied  by  us;  see 
discussion  below];  Allen,  1953,  pp.  308,  311,  315  (in  part)  [Australia,  from 
Xoosa,  Queenland  to  Carnarvon,  "Western  Australia;  name  only;  see  discussion 
below]  ;  Fauvel,  1953,  pp.  474-476,  not  Fig.  249a-o  [India,  Ennur  Backwater; 
description;  most  likely  same  material  as  above,  1932;  see  discussion  below]; 
Rullier,  1955,  pp.  288-289  [Ivory  Coast,  Abidjan;  name  only;  material  from  same 
locality  studied  by  us;  see  discussion  below]  ;  Dew,  1959,  pp.  29-31,  not  Fig.  8A-H 
(in  part)  [Australia,  several  localities;  description;  material  from  Queensland 
(Townsville  and  Noosa)  is  F.  nscliakoz'i,  specimens  from  other  localities  are  F. 
enigmaticus]  ;  Straughan  1966,  pp.  139-146,  Figs.  2,  3b-d  (in  part)  [Australia, 
several  localities;  Brunei;  Ceylon;  some  of  this  material  studied  by  us;  Straughan's 
Figs.  3a  and  3e  are  F.  enigmaticus;  see  discussion  below7]  ;  Rullier,  1966,  pp.  95-104 
(in  part)  [Dahomey,  Cotonou;  other  references  to  F.  uschakovi  in  Rullier's  listing 
are  cited  by  us  in  this  synonymy]  ;  Sandison  and  Hill,  1966,  pp.  235-250  [Nigeria, 
Lagos;  name  only;  see  discussion  below]  ;  Day,  1967,  p.  812  (in  part?)  [South 
Africa,  Natal;  diagnosis;  should  be  checked  since  locality  is  in  tropical  region]; 
Hill,  1967,  pp.  303-321  [Nigeria,  Lagos;  name  only;  see  discussion  below]  ;  Nel- 
son-Smith, 1967,  p.  54  (in  part?)  [the  paleotropical  records  most  probably  are  of 
F.  macrodon  or  F.  itschakovi ;  the  diagnosis  and  figures  are  of  F.  enigmaticus}  ; 
Straughan,  1967,  pp.  25-40  [Australia,  Queensland,  Brisbane  River;  ecological 
study]  ;  Straughan,  1968,  pp.  59-64,  plates  1,  2,  3A  (in  part)  [Australia,  several 
localities;  Straughan's  plate  3B  is  F.  enigmaticus;  see  discussion  below]  ;  Gibbs, 

1971,  p.  203    [Solomon  Islands,  Guadalcanal,  Lunga   Point  and  Komimbo  Bay; 
short  diagnosis;  material  studied  by  us]  ;  Straughan,  1971,  pp.  169-175  [Australia, 
Queensland,  North  Pine  River;  ecological  study;  see  discussion  below]  ;  Straughan, 

1972,  pp.  93-136   [Australia,  Queensland,  Brisbane  River;  ecological  study;  see 
discussion  below7]. 

[  ?]  Mercierella  enigmatica — Day,  1951,  pp.  65-66  [South  Africa,  St.  Lucia  Estuary; 
name  only;  should  be  checked  since  locality  is  in  tropical  region];  Ganapati, 
et  a!.,  1958,  pp.  197-206  [India.  17°  N  83°  E;  name  only;  material  is  probably  F. 
macrodon  or  F.  uschakovi]  ;  Kirkegaard,  1959,  p.  105  [Nigeria,  Lagos,  Victoria 
Beach;  name  only;  see  discussion  below]. 


110  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

non  Mercierella  enigmatica — Mesnil  and   Fauvel,    1939,  pp.  37-38    [Kei   Islands, 

Siboga  Exped.  Stat.  260,  90  m;  one  empty  tube;  see  discussion  below]. 
[?]  Ficopomatus  sp.  Hill,   1967,  pp.  303-321    [Nigeria,  Lagos;  name  only;  most 

likely  abnormal  F.  uschakovi;  see  discussion  below]. 
Neopomatus  uschakovi  Pillai,  1960,  pp.  28-32,  Figs.  10H,  11A-H,  12A-H,  plate 

I,  Figs.  1,  2  [Ceylon,  Panadura  River  Estuary,  Madu  Ganga  Estuary,  Ratgama 
Lake;  description;  holotypes  studied  by  us]  ;  Hartman,  1965,  p.  80  [name  only]  ; 
Pillai,  1965,  p.  172  [Indonesia,  Surabaja;  East  Java;  Madura;  name  only]  ;  Pillai, 
1971,  pp.  118-123,  127,  Figs.  9G,  10A  [Ceylon,  several  localities;  description; 
comparative  study]  ;  Zibrowius,  1973,  p.  64  [synonymy;  useful  discussion]. 
Neopomatus  uschakovi  var.  lingayanensis  Pillai,  1965,  pp.  170-172,  Fig.  23A-I 

[Philippine  Islands,  Luzon,   Lingayan  Gulf  and  other  localities;   description; 
some  paratypes  studied  by  us] . 
Neopomatus  nshakovi   [sic] — Hartmann-Schroder,   1971,  pp.  7-27,  Figs.  2,  3,  5, 

7b-d,  11-14  [several  paleotropical  localities;  partial  revision,  synonymy]. 
Neopomatus  sirnilis  Pillai,  1960,  pp.  32-33,  Fig.  12I-M,  plate  II,  Fig.  1    [Ceylon, 

Negombo  Lagoon;  description,  holotype  studied  by  us]  ;  Hartman,  1965,  p.  80 
[name  only]. 
Neopomatus  sinrilis  var.  rugosus  Pillai,  1960,  pp.  33-35,  plate  II,  Fig.  2  [Ceylon, 

Negombo  Lagoon;  description;  holotype  studied  by  us]  ;  Hartman,  1965,  p.  80 
[name  only]. 

Material  studied.  Sri  Lanka  [--  Ceylon]:  Panadura  River  Estuary,  6  Jan. 
1957  (Holotype  of  N.  uschakovi;  BMNH  1959:  4:  14:7);  Maha  Alamba,  Ne- 
gombo Lagoon,  18  Feb.  1959,  T.  G.  Pillai  coll.  (holotype  of  N.  similis  var.  ntgosus 
and  small  tube  on  a  pebble;  BMNH  1959:  4:  14:  14)  ;  Ratgama  Lake.  28  Feb. 
1959,  coconut  petiole  with  tubes  attached,  T.  G.  Pillai  coll.  (ca.  25  specimens; 
BMNH  1959:  4:  14:  19);  Cuming  coll.,  specific  locality  unknown,  tubes  on 
gastropod  shells  (one  dried  operculum,  empty  tubes;  BMNH  1965:  31:  4-5;  at 
least  110  years  in  BMNH,  identified  by  H.' Zibrowius,  1972).  India:  Madras 
Coast,  Ennur  Backwater,  Annandale  coll.  (four  specimens  in  tubes,  BMNH  1938: 
5:7:  92-94;  also  many  specimens,  some  in  tubes,  MNHN;  as  Tl/.  enigmatica  by 
P.  Fauvel;  as  Neopomatus  sp.  by  G.  Hartmann-Schroder;  as  N.  uschakovi  by  H. 
Zibrowius.  Indonesia.  Java:  Specific  locality  unknown,  1904,  P.  Serre  coll. 
(many  specimens  in  tubes  on  barnacles;  MNHN;  as  M.  enigmatica  by  P.  Fauvel; 
as  TV.  nschakoi'i  by  H.  Zibrowius).  Philippines:  Luzon,  Lingayan  Gulf,  T.  G. 
Pillai  coll.  (five  paratypes  of  TV.  uschakovi  var.  lingayanensis;  BMNH  1965:  53: 
19-28).  Solomon  Islands.  Guadalcanal:  Komimbo  Bay,  19  July  1965,  at  mouth 
of  freshwater  creek,  above  MTL  and  Lunga  Point,  9  Sept.  1965,  in  brackish  lagoon 
at  LWM,  P.  E.  Gibbs  coll.  (ca.  70  specimens,  some  in  tubes;  BMNH  1970:  830/ 
831  ;  as  M.  enigmatica  by  P.  F.  Gibbs,  as  TV.  uschakovi  by  H.  Zibrowius).  Aus- 
tralia. New  South  Wales  :  Yamba,  1  Sept.  1950  and  Queensland  :  Townsville,  26 
Dec.  1950,  and  Noosa,  1  March  1951,  B.  Dew  coll.  (10  specimens;  BMNH  1955: 
11  :  1  :  116;  AM  W-3777-9,  3781  ;  as  M.  enigmatica  by  B.  Dew;  as  N.  uschakovi 
by  H.  Zibrowius  and  T.  G.  Pillai).  Nigeria.  Lagos:  Jan.  1954  (11  specimens  and 
many  others  in  tubes;  BMNH  1954:  3:4:  1-50;  as  Tl/.  enigmatica:  as  Neopomatus 
sp.,  by  G.  Hartmann-Schroder  and  as  TV.  uschakovi  by  H.  Zibrowius).  Ivory 
Coast.  Abidjan :  June  1955,  M.  Fox  coll.  (many  specimens  in  tubes  on  pieces  of 


REVISION  OF  FICOPOMATUS 

wood;  BMXH  1955:  11:1:  1-30;  as  M.  cnii/unitico;  as  Xcopojnafus  sp.  by  Hart- 
mann-Schroder,  and  as  Ar.  uschakot'i  by  H.  Zibrowius.  Netherlands.  Noordwijk: 
on  wood  cast  ashore  on  beach,  13  Oct.  1974,  A.  \Y.  Lacourt  coll.  (many  tubes  and 
dried  opercula,  RMNH  07274,  tHU  213). 

Tube  :  The  tube  is  shining  white,  sometimes  the  older  parts  are  covered  with  a 
brownish  layer  of  algae,  presumably.  It  is  semicircular  in  cross-section.  At  irregu- 
lar intervals  it  bears  more  or  less  prominent  collar-like  rings,  which  indicate  suc- 
cessive positions  of  the  peristome.  Usually  there  are  three  keels  (Fig.  5d),  of 
which  the  median  is  high  and  sharp,  the  lateral  ones  may  be  smaller;  sometimes 
they  are  faint  or  lacking.  The  keels  are  less  conspicuous  toward  the  mouth  of  the 
tube. 

Branchiae :  The  branchial  filaments  arise  from  paired  lobes  and  number  about 
eight  (5-10)  on  the  left  and  nine  (6-11)  on  the  right.  They  are  arranged  in  two 
semicircles  and  are  not  connected  by  a  branchial  membrane.  The  filaments  are 
shorter  ventrally.  The  two  rows  of  pinnulae  become  larger  toward  the  end  of  the 
filaments,  which  is  free  of  pinnulae  to  a  greater  or  lesser  extent  (Fig.  3a). 

Peduncle:  The  peduncle  is  smooth,  sometimes  faintly  wrinkled,  especially  just 
below  the  bulb  of  the  operculum.  It  is  circular  to  subtriangular  (the  latter  near  the 
opercular  bulb)  in  cross-section.  There  is  a  gradual  transition  from  peduncle  to 
opercular  bulb,  however,  slightly  more  abruptly  than  in  F.  cniginaticiis  (cf.  Fig.  2a 
with2f). 

Operculum  :  The  operculum  usually  is  spherical  and  radially  symmetrical,  some- 
times with  bilateral  symmetry.  It  usually  has  a  convex,  slightly  horny  plate 
distally,  which  sometimes  may  be  lacking.  This  end-plate  is  bordered  by  one  to 
four  (exceptionally  up  to  eight)  rows  of  small  denticulations  (Fig.  2a-d),  curved 
outward.  The  denticulations  ("spines"  )  of  one  row  may  be  either  fused  with  or 
completely  separated  from  each  other.  Sometimes  the  rows  of  "spines"  are  incom- 
plete or  irregular.  The  "spines"  are  randomly  placed  in  a  few  specimens,  and, 
exceptionally,  cover  the  endplate.  "Spines"  with  small  outgrowths  sometimes 
occur  (cf.  Fig.  3g,  j  with  h,  i,  k). 

Collar  and  thoracic  membranes :  The  collar  is  rather  high,  not  lobed  and  has  an 
entire  edge.  It  is  continuous  with  the  thoracic  membranes,  which  are  fused 
dorsally  (Fig.  3f)  and  are  united  ventrally  on  the  anterior  abdominal  segments. 
Exceptionally,  there  are  specimens  in  which  the  thoracic  membranes  are  not  fused 
dorsally  (one  of  the  approximately  200  specimens  studied). 

Thorax :  The  thorax  has  seven  segments,  six  of  which  are  uncinigerous.  The 
bundles  of  collar  setae  contain  only  a  few  setae  of  two  types :  coarsely  serrated  ones 
(Fig.  4j-n)  and  limbate  ones.  Subsequent  bundles  of  setae  are  larger  and  are  in 
two  nearly  parallel  rows,  containing  limbate  setae  only  (Fig.  4r).  The  thoracic  un- 
cinigerous tori  are  arranged  in  two  nearly  parallel  rows,  with  up  to  75  uncini 
(n  =  5)  per  torus  in  large  animals  (Lunga  Point).  The  uncini  along  the  entire 
thorax  have  a  single  row  of  seven  to  nine  curved  teeth,  the  most  anterior  tooth 
gouged,  apparently  bifurcated  (Fig.  4x-z). 

Abdomen :  The  number  of  abdominal  segments  is  usually  about  40  (18-46,  n  — 
7).  The  anterior  two  or  three  segments  are  apparently  without  setae  or  uncini. 
The  following  segments  have  very  few  uncini  (three  to  four).  The  number  of 
uncini  per  row  slowly  increases  to  about  45  in  the  middle  of  the  abdomen,  then 


112  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

slowly  decreases  toward  the  pygidium  (three  to  six).  The  abdominal  uncini  are 
rasp-like  along  the  entire  abdomen,  with  two  rows  of  curved  teeth  anteriorly,  two  to 
three  rows  posteriorly;  anteriorly  about  10-12  teeth  are  visible  in  profile,  including 
the  anterior  gouged  one,  posteriorly  about  13  smaller  ones  (Fig.  4jj-mm).  The 
bundles  of  abdominal  setae  consist  of  one  or  two  to  three  geniculate  ones  (Fig.  4yy). 

Size :  The  length,  including  the  operculum,  is  quite  variable.  In  a  population 
from  Lunga  Point  the  length  is  about  10  mm  (6-12  mm)  ;  the  specimens  from 
Komimbo  Bay,  however,  are  not  longer  than  5  mm  (2-5  mm,  n  ==  5).  The  width 
of  the  thorax  is  about  1  mm  in  the  large  specimens,  about  0.4  mm  in  the  small  ones. 
The  branchiae  and  the  operculum  usually  account  for  one-quarter  of  the  entire 
length  of  the  animal. 

Variations :  Special  attention  should  be  given  to  a  form  differing  in  operculum, 
described  by  Pillai  (1965)  as  Neopomatus  uschakovi  var.  lingayanensis.  This 
usually  has  a  bilaterally  symmetrical  operculum,  with  a  cluster  of  one  to  four  spines 
on  the  endplate,  in  the  center  of  the  ring(s)  of  denticulations  (Fig.  2d). 

Discussion:  The  holotype  is  in  poor  condition,  apparently  having  been  dry. 
The  species  has  been  confused  with  F.  cnigniaticus;  however,  Pillai  (1971)  and 
Hartmann-Schroder  (1971)  have  already  clarified  this  confusion  and  indicated 
that  the  species  are  geographically  separated — F.  cnigniaticus  occurs  in  subtropical/ 
temperate  areas,  F.  uschakovi  in  the  paleotropical  region.  The  results  of  our  re- 
search support  this  opinion  (Fig.  6). 

In  eastern  Australia  the  northern  boundary  of  F.  cnigniaticus  and  the  southern 
one  of  F.  uschakovi  lies  just  north  of  Sydney,  according  to  the  material  studied  by 
Pillai  (1971)  and  by  us.  In  western  Australia  it  cannot  as  yet  be  defined  exactly; 
the  population  in  Swan  River  is  F.  cniguiaticus,  the  material  mentioned  by  Allen 
(1953,  p.  308)  from  Carnarvon  might  be  F.  uschakovi,  since  this  locality  lies  within 
the  tropics. 

The  distributions  of  F.  eniguiaticus  and  uschakovi  mdicated  above  suggest  that  it 
is  unlikely  that  both  species  will  occur  together  in  an  entirely  tropical  area.  In 
juvenile  specimens  of  F.  cnigniaticus,  niiainicnsis  and  uschakovi,  opercula  may  have 
no  horny  parts.  Therefore,  Ficopotnatus  sp.,  as  cited  by  Hill  (1967)  from  Nigeria, 
most  likely  is  abnormal  F.  uschakovi  (see  Zibrowius,  1973,  p.  64). 

The  exact  boundaries  between  both  species  in  Africa  cannot  be  given,  since 
there  are  considerable  gaps  in  the  known  distributions. 

Mesnil  and  Fauvel's  (1939,  pp.  37-38)  record  of  an  empty  tube  of  Mercicrclla 
enigmatica  from  a  depth  of  90  m  off  the  Kei  Islands  is  very  doubtful.  Unfortu- 
nately, the  material  could  not  be  traced,  but,  since  many  genera  show  tubes  with 
"peristomes,"  it  is  more  likely  that  this  tube  belonged  to  a  different  genus  than  that 
the  tube  was  deposited  two  miles  offshore  by  streams.  On  the  other  hand,  the  diag- 
nosis and  locality  of  Fillers'  (1918,  p.  250)  record  of  empty  serpulid  tubes  from 
a  river  on  the  Aru  Islands  indicate  that  these  tubes  most  likely  are  F.  uschakovi. 

Our  record  of  F.  uschakovi  from  the  Netherlands  most  probably  can  be  ex- 
plained by  the  brisk  local  trade  in  tropical  wood,  and,  therefore,  has  not  been  in- 
cluded in  Figure  6. 


REVISION  OF  FICOPOMATUS 


113 


114  II.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

J'ii'oponiiiliis  cnii/iinilicns  (Fauvcl,  1923) 
Figures  2e-i ;  3d-e,  1-q;  4a-d,  s,  aa-bb,  nn-vv,  zz  ;  5c 

Due  to  the  large  number  of  literature  citations  of  this  species,  the  following 
represents  a  selected  synonymy  and  is  comprised  of  those  papers  which  have  a 
special  bearing  on  taxonomic  problems  and  those  in  which  material  studied  by  us 
has  been  mentioned.  It  should  be  emphasized  that  Mcrcicrella  cnigniatica  from 
tropical  regions  as  reported  by  Allen  (1953),  Day  (1951,  1967),  Dew  (1959), 
Fauvel  (1931,  1932,  1953),  Ganapati,  ct  al.  (1958),  Gibbs  (1971),  Hill  (1967), 
Kirkegaard  (1959),  Mesnil  and  Fauvel  (1939),  Nelson-Smith  (1967),  Rullier 
(1955,  1966),  Sandison  and  Hill  (1966),  Straughan  (1966,  1967,  1968,  1971, 
1972),  at  least  partly,  belong  to  one  of  the  other  three  species  and  can  be  found  in 
their  respective  synonymies  above. 
Mcrcierella  cnigniatica  Fauvel,  1923,  pp.  424-430,  Fig.  la-o  [France,  Canal  de 

Caen;  description;  syntypes  studied  by  us];  Fauvel,  1931,  pp.  1068-1069  (in 
part)  [several  localities;  name  only;  the  record  of  India,  Ennur  Backwater  is 
probably  F.  uschakovi  or  F.  macrodon]  ;  Rioja,  1931,  pp.  420-424,  plates  137-139 
[Spain,  Gandia;  extensive  description;  account  of  infraspecific  variability]  Fauvel, 
1933,  pp.  185-193  [several  localities;  short  description];  Monro,  1938a,  pp.  311 
and  313  [Uruguay,  Arroyo  de  las  Brujas,  Canelones;  name  only;  material  studied 
by  us];  Monro,  1938b,  p.  624  [Western  Australia,  Pelican,  Swan  River;  name 
only;  part  of  this  material  studied  by  us]  ;  Hartman,  1952,  p.  64  [Texas,  Rockport ; 
diagnosis;  material  studied  by  us]  ;  Allen,  1953,  pp.  308,  311,  and  315  (in  part) 
[Australia,  from  Noosa,  Queensland,  to  Carnarvon,  Western  Australia;  name  only; 
see  discussion  of  F.  uschakovi  above]  ;  Cognetti  1953,  pp.  36-40,  Fig.  la-n  [Italy, 
Toscana;  variability  of  operculum]  ;  Day,  1955,  p.  448  [South  Africa,  several 
localities  on  the  Cape;  name  only;  part  of  this  material  studied  by  us]  ;  Dew,  1959, 
pp.  29-31,  Fig.  SA-H  (in  part)  [Australia,  several  localities;  the  material  from 
Queensland  (Townsville  and  Noosa)  is  F.  iischakovi;  specimens  from  other 
localities  are  F.  cniyiinaticits;  description ;  part  of  this  material  studied  by  us]  ; 
Hartman,  1959,  p.  582  [name  only]  ;  Pillai,  1960,  p.  33  [comparison  with  other 
species]  ;  Yuillemin,  1964,  pp.  514-527,  plates  1-5  [Tunisia,  Lac  de  Tunis;  exten- 
sive description  of  opercular  variability]  ;  Yuillemin,  1965,  554  pages,  many  figures 
[Tunisia,  Lac  de  Tunis;  thesis  on  their  biology]  ;  Hartman,  1966,  p.  238  [Hawaii, 
Oahu,  Honolulu,  Alai  Wai  Canal  near  Waikiki  ;  diagnosis;  material  from  same 
locality  studied  by  us]  ;  Rullier,  1966,  pp.  95-104  (in  part)  [list  of  localities  to  1964, 
from  the  literature;  tropical  records  probably  are  F.  uschakovi]  ;  Straughan,  1966, 
pp.  139-146,  Fig.  3a,  e  (in  part)  [Australia,  several  localities,  and  California,  Ber- 
keley; specimens  in  Figs.  2  and  3b-d  are  F.  uschakori;  see  discussion  above]  ;  Day, 
1967,  p.  812  (in  part)  [South  Africa,  several  localities  on  the  Cape;  diagnosis;  the 
record  from  Natal  (tropical)  should  be  checked  for  it  may  be  F.  nschakoi'i]  ;  Hart- 
mann-Schroder,  1967,  pp.  421-456,  Figs.  1-24  [Europe,  several  localities;  mono- 
graph of  the  species]  ;  Nelson-Smith,  1967,  p.  54,  Figs.  49-50  [Southwestern 
United  Kingdom  ;  diagnosis ;  tropical  localities  in  the  distribution  most  probably 
represent  other  species,  see  discussions  above]  ;  Straughan,  1968,  pp.  59-64,  plate 
3B  (in  part)  [Australia,  several  localities;  the  material  on  plates  1,  2,  and  3A  are 
F.  uschakovi;  see  discussion  above]  ;  Wolff,  1969,  pp.  85-92,  Figs.  1-6  [Southwest- 


REVISION  OF  FICOPOMATUS 

ern  Netherlands;  extensive  description;  part  of  material  studied  by  us]  ;  Hartmann- 
Schroder,  1971,  pp.  7-27,  Figs.  1,  4,  6,  7a,  8-10,  15-17  [Mediterranean,  several 
localities,  Black  Sea,  and  Australia,  New  South  \Yales;  partial  revision  and 
synonymy]  ;  Orensanz  and  Estivariz,  1971,  pp.  106-108,  Figs.  47-56  [Argentina, 
several  localities;  diagnosis;  part  of  this  material  studied  by  us];  Pillai, 
1971,  pp.  120-125,  Fig.  10B-H  [United  Kingdom,  Radipole  Lake,  Weymouth; 
description;  comparative  study  of  the  four  species];  Zibrowius,  1973,  pp.  62-64 
[useful  discussion]  ;  Hove,  ten  1974,  pp.  45-48  [Southwestern  Netherlands;  name 
only;  material  studied  by  us];  Kajihara,  Hirano  and  Chiba,  1976,  pp.  363-366 
[Japan,  Hamana-ko ;  name  only];  Bailey-Brock,  1976,  p.  73  [Hawaii,  Oahu, 
several  localities;  name  only;  part  of  material  studied  by  us]. 

Material  studied.  France :  Canal  de  Caen,  19  Sept.  1922,  L.  Mercier  coll. 
(three  syntypes;  BMNH  1928:  4 :  26 :  16-17).  Netherlands:  Vlissingen,  inner 
harbor,  L.  de  Wolf  coll.  (empty  tubes;  tHU  78,  ZMU;  as  M.  enigmatica  by  W.  J. 
Wolff)  ;  Ylissingen,  Keersluisburg,  6  April  1972  and  25  Sept.  1973,  on  piling  near 
power  station,  about  1  m  deep.  H.A.  ten  Hove  coll.  (very  many  specimens,  tHU 
169,  191,  ZMA  V.  Pol.  2615,  SME).  Tunisia:  Lac  de  Tunis,  1969,  B.  Hotman 
coll.  (many  specimens;  tHU  85;  as  Jl/.  enigmatica  by  H.  Zibrowius).  South 
Africa:  Cape  Town,  Milnerton  Estuary  (two  specimens,  20  tubes;  BMNH  1952: 
8:  10:  1 ;  as  HI.  enigmatica  by  J.H.  Day).  Australia:  Western  Australia,  Pelican, 
Swan  River.  17  Aug.  1935,  D.  L.  Serventy  coll.  (two  specimens  and  others  in  tubes; 
BMNH  1938:  10:  31:  29-32;  as  M.  'enigmatica  by  C.  C.  A.  Monro  and  H. 
Zibrowius)  ;  New  South  Wales,  Sydney,  Tempe,  Cooke's  River,  B.  Dew  coll. 
(fragmentary  specimen,  clusters  of  tubes ;  BMNH  1955 :  9 :  2 :  1-20 ;  as  M. 
enigmatica  by  N.  Tebble).  United  States  of  America:  Texas,  Rockport,  28  Sept. 
1951,  fouling  on  bottom  of  boat,  G.  Gunter  coll.  (two  specimens;  AHF)  ;  Cali- 
fornia, Oakland,  Lake  Merritt,  1931  (ten  specimens,  20  tubes;  tHU  232;  as  M. 
enigmatica  by  P.  Fauvel)  ;  Hawaii,  Oahu,  Honolulu,  Alai  Wai  Canal  near  Waikiki, 
J.  H.  Bailey-Brock  coll.  (three  specimens  and  several  others  in  tubes;  tHU  163;  as 
M.  enigmatica  by  J.  H.  Bailey-Brock).  Uruguay:  Las  Brujas,  Canelones,  25  July 
1937  (many  specimens  in  tubes;  BMNH  1937:  10:  15:  1-10;  as  M.  enigmatica  by 
C.  C.  A.  Monro  and  H.  Zibrowius).  Argentina:  Buenos  Aires  (prov.),  Albufera 
de  Mar  Chiquita,  desembocadura  del  Canal  7,  12  Oct.  1968,  J.  M.  Orensanz  coll. 
(32  specimens ;  tHU  150). 

Tube :  The  tube  is  white,  sometimes  covered  with  a  brown  layer,  presumably 
algae.  It  is  semicircular  to  circular  in  cross-section.  At  irregular  intervals  it  often 
bears  wide,  flaring,  sometimes  faint,  collar-like  rings  indicating  the  successive  posi- 
tions of  the  peristome  (Fig.  5c).  Solitary  or  juvenile  tubes  sometimes  have  a 
faint  median  keel  (see  Cognetti,  1954,  Fig.  1). 

Branchiae :  The  branchial  filaments  arise  from  paired  lobes  and  number  about 
seven  (5-9)  on  the  left  and  eight  (7-10)  on  the  right.  They  are  arranged  in  two 
semicircles  and  are  not  connected  by  a  branchial  membrane.  The  filaments  are 
somewhat  shorter  ventrally.  The  two  rows  of  pinnulae  become  larger  towards  the 
ends  of  the  filaments,  which  are  free  of  pinnulae  to  a  greater  or  lesser  extent  (Fig. 
3d-e). 

Peduncle :  The  peduncle  is  smooth,  sometimes  faintly  wrinkled,  especially 
below  the  bulb  of  the  operculum  (Fig.  2g)  ;  it  is  subtriangular  in  cross-section  with 


116  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

a  shallow  dorsal  groove  (Fig.  2f).  There  is  a  gradual  transition  between  peduncle 
and  opercular  bulb  (Fig.  2f). 

Operculum  :  The  operculum  is  fig-shaped,  usually  bilaterally  symmetrical  with  a 
distal  eccentrically  placed  concavity.  The  concave  part  generally  has  a  horny  plate, 
bordered  by  one  to  four  rows  of  spines,  curved  inward  (Fig.  2f-h).  The  rows  of 
spines  may  be  incomplete  dorsally,  or  somewhat  irregular  (Fig.  2h).  The  spines 
are  randomly  placed  in  a  few  specimens  (Fig.  2i).  Exceptionally  the  operculum 
lacks  spines  (Fig.  2e).  The  spines  sometimes  have  one  to  three  short  radial  ac- 
cessory spines  (Fig.  31-q). 

Collar  and  thoracic  membranes :  The  collar  is  high,  not  lobed  and  has  an  entire 
edge.  It  is  continuous  with  the  thoracic  membranes  which  are  united  ventrally  on 
the  anterior  abdominal  segments. 

Thorax :  The  thorax  has  seven  segments,  six  of  which  are  uncinigerous.  The 
collar  setae  are  of  two  types:  coarsely  serrated  (Fig.  4a-d)  and  limbate.  Subse- 
quent bundles  of  setae  are  larger  and  are  in  two  nearly  parallel  rows,  containing 
limbate  setae  only  (Fig.  4s).  The  thoracic  uncinigerous  tori  are  arranged  in  two 
nearly  parallel  rows,  with  up  to  90  uncini  per  torus.  The  uncini  along  the  entire 
thorax  have  a  single  row  of  six  to  seven  curved  teeth ;  the  most  anterior  tooth  is 
gouged,  apparently  bifurcated  (Fig.  4aa-bb). 

Abdomen:  The  number  of  abdominal  segments  is  usually  about  60  (29-84; 
n  =  7).  The  anterior  two  or  three  segments  are  apparently  without  setae  or  un- 
cini. The  following  segments  have  relatively  few  uncini  (21-35),  the  number  per 
row  increasing  rapidly  in  the  anterior  one-third  of  the  abdomen  (80-120),  then 
slowly  decreasing  towards  the  pygidium  (3-20).  The  abdominal  uncini  of  the 
anterior  segments  have  a  single  row  of  curved  teeth  (six  to  seven),  including  the 
anterior  gouged  tooth;  the  uncini  of  the  posterior  segments  are  smaller,  with  two 
rows  of  small  curved  teeth,  with  10-12  teeth  visible  in  profile,  including  the  anterior 
gouged  one  (Fig.  4nn-vv).  The  bundles  of  abdominal  setae  consist  of  two  to  five 
geniculate  ones  (Fig.  4zz). 

Size:  The  length,  including  the  operculum,  is  usually  about  20  mm  (7-44). 
The  width  of  the  thorax  is  about  1  mm  (0.9-1.2).  The  branchiae  and  the  opercu- 
lum usually  account  for  one-sixth  of  the  entire  length  of  the  animal. 

Discussion:  Ficopomatus  enigmaticus  is  mentioned  in  well  over  150  papers,  in 
various  fields  of  research.  We  want  to  emphasize  that  some  important  ecological 
works  have  been  based  upon  incorrectly  identified  material.  Therefore,  the  results 
of  this  research  can  be  evaluated  only  after  a  careful  comparison  with  the  distribu- 
tional data,  given  in  this  paper  (Fig.  6). 

Records  from  Japan  (Okayama  Pref.,  Kujima  Lake;  Tokyo,  Sumida  River; 
Ryukyu  Islands,  Ishigaki-jima,  Kabin  Bay)  have  been  confirmed  by  an  excellent 
unpublished  figure  by  M.  Imajima. 


The  authors  wish  to  express  their  thanks  for  the  loans  or  donations  of  material 
to  Dr.  Julie  H.  Bailey-Brock,  University  of  Hawaii,  Honolulu ;  Dr.  K.  Fauchald 
(AHF)  ;  Dr.  J.  D.  George  (BMNH)  ;  Dr.  Pat  Hutchings  (AM)  ;  Dr.  M.  L. 
Jones  and  Dr.  Marian  H.  Pettibone  (USNM)  ;  Dr.  E.  Kirsteuer  (AMNH)  ;  Dr. 


REVISION  OF  F1COPOMATUS  1 < 

T.  Lacalli,  Huntsman  Marine  Laboratory,  St.  Andrews,  Canada;  Dr.  J.  van  drr 
Land  (RMNH)  ;  Dr.  J.  M.  Orensanz,  Institute  Biologia  Marina,  Playa  Grande, 
Argentina;  Dr.  I.  Renaud-Mornant  (MXHN)  ;  Dr.  F.  Rullier,  Universite  Catho- 
lique  de  1'Ouest/ Angers ;  Dr.  S.  van  der  Spoel  (ZMA)  ;  Dr.  B.  A.  Vittor  and  Mr. 
P.  G.  Johnson,  Dauphin  Island  Sea  Laboratory,  Alabama;  Dr.  P.  \Yagenaar  Hum- 
melinck  (ZMU)  ;  Dr.  \V.  ].  Wolff,  Delta  Instituut,  Yerseke;  and  Dr.  H.  Zibrowius 
(SME). 

Thanks  are  also  due  to  Dr.  M.  Imajima,  National  Science  Museum,  Tokyo, 
for  drawing  attention  to  the  occurrence  of  F.  cniginaticits  in  Japan,  and  for  per- 
mission to  include  his  unpublished  distributional  data  in  this  paper.  A  grant  of  the 
Netherlands  Foundation  for  the  Advancement  of  Tropical  Research  (WOTRO) 
enabled  the  senior  author  to  collect  and  study  living  specimens  in  the  Netherlands 
Antilles.  Dr.  J.  D.  George  (BMXH),  Dr.  M.  L.  Jones,  and  Dr.  M.  H.  Pettibone 
(USNM)  kindly  read  the  manuscript  critically.  The  authors  are  responsible  for 
the  remaining  faults. 

SUMMARY 

The  brackish  water  serpulid  genera  Mercicrclla,  Merrier  ellopsis,  Neopomatus 
and  Sphacropouiatus  are  synonymizecl  with  Ficopomatus,  including  four  species: 
F.  enigmatic  us,  F.  macrodon,  F.  miamiensis  and  F.  iisclwkori.  The  geographical 
distributions  of  the  species  are  illustrated,  and  the  confused  identity  of  tropical 
specimens  has  been  clarified,  at  least  in  part.  The  generic  position  of  Ficopomatus 
capensis  is  discussed.  Fossil  records  of  Mercicrclla  and  related  genera  most  prob- 
ably do  not  belong  to  the  genus  Ficopomatus. 

LITERATURE  CITED 

ALLEN,  F.  E.,  1953.     Distribution  of  marine  invertebrates  by  ships.     Aust.  J.  Mar.  Freshwater 

Res.,  4:307-316. 
ANNANDALE,  N.,  1916.     Preliminary  report  on  the  fauna  of  the  Tale  Sap  or  inland  sea  of  Sing- 

gora.    /.  Nat.  Hist.  Soc.  Siam,  2:  90-102. 
BAILEY-BROCK,  J.  H.,  1976.    Habitats  of  tubicolous  polychaetes  from  the  Hawaiian  Islands  and 

Johnston  Atoll.  Pacific  Sci.,  30:  69-81. 
COGNETTI,  G.,  1953.     Variazoni  delle  spine  opercolari  in  nuove  stazioni  di  Mercicrclla  cnigma- 

tica  Fauvel  (Polychaeta,  Serpulidae)   sulla  costa  toscana.     Proccssi  Verb.  Soc.  Toscana 

Sci.  Nat.  Pisa  Ser.  B.,  60 :  36-40. 
COGNETTI,  G.,  1954.     Forme  della  Mcrcicrella  cniginatica  Fauvel  nella  nuova  stazione  del  Lago 

di  Patria.    Boll.  Zool.,  21 :  41-44. 
DAY,  J.  H.,  1951.     The  polychaet  fauna  of  South  Africa.    Pt.  1.     The  intertidal  and  estuarine 

Polychaeta  of  Natal  and  Mosambique.    Ann.  Natal  Mus..  12(1)  :  1-67. 
DAY,   J.   H.,   1955.     The  Polychaeta   of   South  Africa.     Pt.   3.      Sedentary   species   from   Cape 

shores  and  estuaries.    /.  Linn.  Soc.  Land.  Zool.,  42:  407-452. 
DAY,  J.  H.,  1961.     The  polychaet  fauna  of  South  Africa.     Pt.  6.     Sedentary  species  dredged 

off  Cape  coasts  with  a  few  new  records  from  the  shore.    /.  Linn.  Soc.  Land.  Zool.,  44: 

463-560. 
DAY,  J.  H.,  1967.     A  monograph  on  the  Polychaeta  of  Southern  Africa,   Part  2.    Sedentaria. 

Publ.  Brit.  Mus.  (Nat.  Hist.)  London,  656:    459-878. 

DEW,  B.,  1959.    Serpulidae  (Polychaeta)  from  Australia.    Rec.  Aust.  Mus..  25:  19-56. 
DRAGASTAN,  O.,  1966.     A  new  serpulid  species  in  the  Upper  Jurassic  of  Rumania.    Palaeontol. 

Z.,  40:  147-150. 
EHLERS,    E.,    1918.      Polychaete    Anneliden    von    den    Am-    und    Kei-Inseln.    Abh.    Scnckcnb 

Naturforsch.  Gcs.,  35  :  227-250. 


118  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

FAUVEL,   P.,   1923.     Un   nouveau   scrpulicn   dVau   saumatre   Mcrcierella   n.g.   enigmatica  n.   sp. 

'  Bull.  Soc.  Zool.  Fr.,  46  :  424-430. 

FAUVEL,  P.,  1931.    Les  migrations  d'une  annelide.    Bull.  Soc.  Occanogr.    Fr.,  11:  1067-1069. 
FAUVEL,  P.,  1932.     Annelida  Polychaeta  of  the  Indian  Museum,  Calcutta.     Mem.  Indian  Mus., 

'  12(1)  :  1-262. 
FAUVEL,   P.,    1933.     Histoire  de   la  Mcrcierella   enigmatica   Fauvel,   serpulien   d'eau   saumatre. 

.  Irch.  Zool.  Ex  p.  Gen.,  75 :  185-193. 
FAUVEL,  P.,  1953.     The  fauna  of  India,  including  Pakistan,  Ceylon,  Burma  and  Malaya.    The 

Indian  Press,  Allahabad,  507  pp. 
GANAPATI,   P.   N.,   M.   V.   LAKSHMANA   RAO,   AND   R.   NAGABHUSHANAM,    1958.     Biology   of 

fouling  in  the  Visakhapatnam  Harbour.     Andhra  Univ.  Mem.  Oceanogr.,  2(62)  :   193- 

209. 
GIBBS,   P.  E.,   1971.     The  polychaete  fauna  of  the   Solomon   Islands.     Bull.   Brit.  Mus.    (Nat. 

Hist.)  Zool. ,21:  99-211. 
HARTMAN,  O.,  1951.    The  littoral  marine  annelids  of  the  Gulf  of  Mexico.   Publ.  Inst.  Mar.  Sci. 

Univ.  Texas,  2  :  7-124. 
HARTMAN,  O.,  1952.     Fouling  serpulid  worms,  new  to  the  Gulf  of  Mexico.    Texas  J.  Sci.,  4: 

63-64. 
HARTMAN,  O.,  1954.     Polychaetous  annelids  of  the  Gulf  of  Mexico.    U.  S.  Fish.  Wildl.  Serv. 

Fish.  Bull,  55:  413-417. 
HARTMAN,  O.,  1956.     Polychaetous  annelids  erected  by  Treadwell  1891-1948,  together  with  a 

brief  chronology.     Bull.  Am.  Mus.  Nat.  Hist.,  109:  239-310. 
HARTMAN,  O.,  1959.     Catalogue  of  the  Polychaetous  Annelids  of  the  World.  Part  II.    Allan 

Hancock  Found.  Publ.  Occas.  Pap.,  23  :  354-628. 
HARTMAN,  O.,  1965.     Catalogue  of  the  Polychaetous  Annelids  of  the  World.    Supplement  1960- 

1965  and  Index.    Allan  Hancock  Found.  Publ.  Occas.  Pap.,  23   (Suppl.)  :  1-197. 
HARTMAN,   O.,   1966.     Polychaetous   annelids   of  the   Hawaiian   Islands.      Occas.   Pap.   Bernice 

Pauahi  Bishop  Mus.,  23  :  163-252. 
HARTMANN-^CHRODER,    G.,    1967.     Zur   Morphologic,   Okologie   und   Biologic   von   Mcrcierella 

enigmatica  (Serpulidae,  Polychaeta)   und  ihrer  Rohre.     Zool.  Ans.,  179:  421-456. 
HARTMANN-SCHRODER,  G.,  1971.     Zur  Unterscheidung  von  Neopomatus  Pillai  und  Mcricerella 

Fauvel    (Serpulidae,   Polychaeta).      (Mit  neuen   Beitragen  zur  Kenntnis  der  Okologie 

und  der  Rohrenform  von  Mcrcicrella  enigmatica  Fauvel).    Mitt.  Hamburg  Zool.  Mus. 

Inst.,  67  :  7-27. 
HILL,  M.  B.,  1967.    The  life  cycles  and  salinity  tolerance  of  the  serpulids  Mcrcierella  enigmatica 

Fauvel  and  Hydroidcs  uucinata  (Philippi)  at  Lagos,  Nigeria.    /.  Anim.  Ecol.,  36:  303- 

321. 
HOVE,  H.  A.  TEN,   1970.     Serpulinae   (Polychaeta)    from  the  Caribbean:    I — The  genus  Spiro- 

branchus.  Stud.  Fauna  Curasao,  32  :  1-57. 
HOVE,  H.  A.  TEN,  1973.     Serpulinae  (Polychaeta)   from  the  Caribbean:    II — The  genus  Sclero- 

styla.    Stud.  Fauna  Curasao,  43  :  1-21. 

HOVE,  H.  A.  TEN,  1974.     Notes  on  Hydroidcs  elcgans  (Haswell,  1883)   and  Mcrcierella  enigma- 
tica  Fauvel,   1923,   alien   serpulid   polychaetes   introduced   into   the   Netherlands.     Bull. 

Zool.  Mus.  Amsterdam,  4:  45-51. 
HOVE,    H.    A.    TEN,    1975.      Serpulinae    (Polychaeta)    from    the    Caribbean:     III — The    genus 

Pscudoz'crmilia  (including  species  from  other  regions).    Stud.  Fauna  Curacao,  47:  46- 

101. 
HOVE,  H.  A.  TEN,  1978.     Different  causes  of  mass  occurrence  in  serpulids.  Syst.  Assoc.  Spec. 

Vol.,  in  press. 
KAJIHARA,   T.,    R.    HIRANO,    AND   K.    CHIBA,    1976.      Marine   fouling   animals    in    the    Bay   of 

Hamana-ko,  Japan.   Veligcr,  18 :  361-366. 
KIRKEGAARD,  J.  B.,  1959.    The  Polychaeta  of  West-Africa.    Pt.  I.  Sedentary  species.  Atl.  Rep., 

5:7-117. 
LACALLI,  T.,  1977.     Remarks  on  the  larvae  of  two  serpulids   (Polychaeta)   from  Barbados.  Can. 

J.  Zool,  55  :  300-303. 
MclNTOSH,  W.  C.,  1885.    Report  on  the  Annelida  Polychaeta  collected  by  H.M.S.  CHALLENGER 

during  the  years  1873-1876.     Rep.  Sci.  Results  Challenger  (Zool.),  12:  1-554. 
MclNTOSH,  W.  C.,  1926.     Notes  from  the  Gatty  Marine  Laboratory,  St.  Andrews.  No.  XLIX. 


REVISION  OF  FICOPOMATUS 

1.    On  the  structure  and  functions  of  the  operculum  and  neighbouring  parts  of  Mer- 

cierclla  cnigmatica  Fauvel,  and  other  serpulids.    Ann.  Mag.  .\'at.  Hist.  (9),  18:  402-424. 

MESNIL,  F.,  AND  P.  FAUVEL,  1939.     Polychetes  sedentaires  de  1'expedition  du  "Siboga".     Mal- 

danidae,   Cirratulidae,   Capitellidae,   Sabellidae  et   Serpulidae.    Siboga  Expcd.,  24   (2)  : 

1-42. 

MONRO,  C.  C.  A.,  1938a.     On  a  small  collection  of  Polychaeta  from  Uruguay.     Ann.  Mag.  Nat. 

Hist.  (11),  2:  311-314. 
MONRO,    C.   C.   A.,    1938b.    On   a   small   collection   of   Polychaeta   from    Swan    River,   Western 

Australia.     Ann.  Mag.  Nat.  Hist   (11),  2:  614-624. 
MORCH,  O.  A.  L.,  1863.     Revisio  Critica  Serpulidarum.    Etbidrag  til  rpYormenes  naturhistorie. 

Naturh.   Tidsk.  Hcnrik  Kr0ycr,  Kpbenhavn    (3),   1:   347-470. 
NELSON-SMITH,   A.,    1967.      Catalogue  des  principals  salissures  marines.      Vol.   III.   Scr pules 

titbicolcs.    Organization  for  Economic  Co-operation  and  Development,  Paris,  79  pp. 
ORENSANZ,  J.  M.,  AND  M.  C.  ESTIVARIZ,  1971.    Los  anelidos  poliquetos  de  aguas  salobres  de  la 

Provincia  de  Buenos  Aires.    Rev.  Mus.  La  Plata   (Secc.}   Zoo!.,  11:  95-114. 
PILLAI,  T.  G.,  1960.     Some  marine  and  brackish-water  serpulid  Polychaeta  from  Ceylon,  includ- 
ing new  genera  and  species.     Ceylon  J.  Sci.  Biol.  Sci.,  3  :   1-40. 
PILLAI,  T.  G.,  1965.     Annelida  Polychaeta  from  the  Philippines  and  Indonesia.    Ceylon  J.  Sci. 

Biol.  Sci.,  5:  110-177. 
PILLAI,  T.  G.,  1971.     Studies  on  a  collection  of  marine  and  brackish-water  polychaete  annelids 

of  the  family  Serpulidae  from  Ceylon.     Ceylon  J.  Sci.  Biol.  Sci.,  9:  88-130. 
REGENHARDT,    H.,    1961.      Serpulidae    (Polychaeta   sedentaria)    aus   der   Kreide    Mitteleuropas, 

ihre  okologische,  taxionomische  und  stratigraphische  Bewertung.    Mitt.  Geol.  Staatsinst. 

Hamb.,30:  5-115. 
RIOJA,   E.,    1924.     La  Mercierella   cnigmatica    Fauvel,   serpulido   de   agua   salobre,    en   Espana. 

Bol.  R.  Soc.  Esp.  Hist.  Nat.  Madrid,  24:  160-169. 
RIOJA,  E.,  1931.    Estudio  de  los  poliquetos  de  la  Peninsula  Iberica.     Mem.  Acad.  Cicnc.  Exact. 

Fis.  Nat.  Madrid,  2:  1-471. 
RIOJA,  E.,  1945.     Estudios  anelidologicos  XIII.     Un  nuevo  genero  de  serpulido  de  agua  salobre 

de  Mexico.    An.  hist.  Biol.  Mexico.  16:  411-417. 
RULLIER,    F.,    1955.      Station   nouvelle   de   Mercierella   cnigmatica    sur   la    Cote    d'lvoire.     Vie 

Milieu,  6 :  288-289. 
RULLIER,    F.,    1966.      La   propagation    de    Mercierella    cnigmatica    Fauvel    (annelide    polychete 

sedentaire)  dans  le  monde  entier,  en  moins  de  cinquante  ans.  Mem.  Soc.  Nat!.  Sci.  ATat. 

Math.  Cherbourg,  51 :  95-104. 
SANDISON,  E.  E.,  AND  M.  B.  HILL,  1966.     The  distribution  of  Balanus  pallidus  stutsburi  Darwin, 

Gryphaca  gasar    ((Adanson)    Dautzenberg),   Mercierella   cnigmatica   Fauvel   and  Hy- 

droidcs   uncinata    (Philippi)    in    relation   to   salinity   in    Lagos    Harbour   and   adjacent 

creeks.  /.  Anim.  Ecol.,  35:  235-250. 
SCHMIDT,  W.  J.,  1951.    Neue  Serpulidae  aus  dem  tertiaren  Wiener  Becken.   Ann.  Naturh.  Mus. 

U'icn  Gcol.  Palacontol.,  58:  77-84. 
SOUTHERN,  R.,  1921.     Polychaeta  of  the  Chilka  Lake  and  also  of  fresh  and  brackish  waters  in 

other  parts  of  India.     Mem.  Indian  Mus.,  5:    563-659. 
STRAUGHAN,   D.,    1966.     Australian   brackish   water   serpulids    (Annelida:     Polychaeta).     Rec. 

Aust.Mus.,27:  139-146. 
STRAUGHAN,   D.,   1967.     Intertidal   fouling    in   the   Brisbane   River,   Queensland.   Proc.   R.   Soc. 

Queensland,  79:  25-40. 

STRAUGHAN,  D.,  1968.    Ecological  aspects  of  serpulid  fouling.    Aust.  Nat.  Hist.,  16:  59-64. 
STRAUGHAN,   D.,    1971.     Establishment   of   non-breeding   population   of   Mercierella    enigmatica 

(Annelida:    Polychaeta)   upstream  from  a  breeding  population.     Bull.  S.  C.  Acad.  Sci 

69:  169-175. 

STRAUGHAN,  D.,  1972.     Ecological  studies  of  Mercierella  enigmatica  Fauvel   (Annelida:    Poly- 
chaeta)  in  the  Brisbane  River.    /.  Anim.  Ecol.,  41 :  93-136. 
TEN  HOVE,  H.  A.   See  Hove,  H.  A.  ten. 

TREADWELL,  A.  L.,  1934.    Sphacropmnatus  miamiensis,  a  new  genus  and  species  of  serpulid  poly- 
chaete.   /.  Wash.  Acad.  Sci..  24:  338-341. 
VUILLEMIN,  S.,  1964.     Polymorphisme  operculaire  du  serpulien  Mercierella  cnigmatica  Fauvel 

(Annelide  polychete).     Bull.  Soc.  Zoo/.  Fr.,  89:  514-527. 


120  H.  A.  TEN  HOVE  AND  J.  C.  A.  WEERDENBURG 

VUILLEMIN,  S.,  1965.    Contribution  a  1'etude  ecologique  de  lac  de  Tunis.   Biologic  de  McrclcrcUa 

cmgmatica  Fauvcl.     These,  Univ.  Paris  (A)  4622,  5469,  554  pp. 
WOLFF,  W.  J.,  1969.     Mercierella  enigmatica  Fauvcl,  ccn  borstelworm  van  hot  brakke  water, 

voor  het  eerst  in  Nedorland  gcvondcn.    Lcvcnde  Nat.,  72 :  85-92. 
ZIBROWIUS,  H.,   1969.     Review  of  some  little  known  genera  of   Serpulidac.     (Annelida:   Poly- 

chaeta).    Smithson.  Contrib.  Zool.,  42:  1-22. 
ZIBROWIUS,  H.,   1973.     Serpulidae   (Annelida  Polychaeta)    des  cotes  ouest  de  1'Afrique  et  des 

archipels  voisins.   Ann.  Mus.  R.  Afr.  Cent.  Scr.  Quart.  Zool.,  207 :  1-93. 


Reference:  Biol.  Bull,  154:  121-137.     (February,  1978) 


LIFE  CYCLE,  DISTRIBUTION  AND  ABUNDANCE  OF  CARCINONE- 

MERTES  EPIALTI,  A  NEMERTEAN  EGG  PREDATOR  OF  THE 

SHORE  CRAB,  HEMIGRAPSUS  OREGONENSIS,  IN 

RELATION  TO  HOST  SIZE,  REPRODUCTION 

AND  MOLT  CYCLE 

ARM  AND  M.  KURIS 
Department  of  Biological  Sciences,  University  of  California,  Santa  Barbara,  California  93106 

The  conceptualization  of  the  host  as  a  microenvironment  for  the  symbiont 
(Pavlovski,  1934)  provided  the  seed  for  the  growth  of  the  field  of  parasite  ecology. 
However,  the  difficulties  involved  in  quantifying  the  biology  of  two  disparate 
organisms,  host  and  parasite,  have  impeded  the  study  of  host-symbiont  systems 
from  an  ecological  perspective.  As  hosts,  arthropods  lend  themselves  well  to 
symbiont  population  studies.  Cyclical  events  through  the  course  of  successive  molt 
cycles  impose  many  restrictions  on  aspects  of  host  growth  and  reproduction.  Thus, 
many  host  life  history  events  are  discrete  and  amenable  to  quantification. 

In  this  study  populations  of  the  nemertean  egg  predator,  Carcinonemertes 
epialti  Coe,  1902  were  monitored  in  relation  to  the  biology  of  its  host,  Hemigrapsus 
oregonensis  (Dana,  1851),  a  common  intertidal  shore  crab  along  the  west  coast 
of  North  America.  Adult  specimens  of  C.  epialti  are  only  found  within  or  adjacent 
to  egg  masses  of  female  crabs.  They  live  in  mucoid  tubes  of  their  own  construc- 
tion. Since  C.  epialti  adults  feed  on  crab  eggs,  host  reproductive  conditions  are  a 
primary  factor  in  the  worm's  biology. 

The  nonfeeding  juvenile  portion  of  the  nemertean's  life  cycle  is  spent  en- 
sheathed  on  the  exoskeleton  of  host  crabs  of  either  sex.  Newly  molted  crabs  lack 
nemerteans.  The  nemertean  population  of  the  previous  instar  is  shed  at  ecdysis. 
Nemertean  transmission  through  a  host  molt  cycle  should  result  in  increasing 
nemertean  density  on  the  host  with  advancing  stages  in  the  molt  cycle.  Further- 
more, crabs  with  longer  intermolt  cycles,  such  as  large  crabs,  should  tend  to  have 
more  nemerteans.  Thus,  C.  epialti  is  regarded  as  a  population  of  animals  dis- 
seminating through  a  habitat  which  consists  of  systematically  renewed  substrates, 
crab  exoskeletons.  Access  to  the  host  cuticle  can  be  partially  estimated  by  a  size 
and  sex-specific  determination  of  the  host's  molt  stage.  Several  studies  have  com- 
pared the  biology  of  epizoic  organisms  with  the  general  pattern  of  the  host  molt 
cycle.  Barnacles  have  been  used  to  estimate  the  molting  frequency  of  lobster 
(Nephrops)  hosts  (Barnes  and  Bagenal,  1951).  The  barnacle  Trilasmis  repro- 
duces more  frequently  than  the  average  spiny  lobster  host  intermolt  duration,  as- 
suring continual  replenishment  of  epizoic  populations  (Bowers,  1968).  Apostome 
ciliates  excyst  and  initiate  feeding  at  host  ecdysis  (Trager,  1957;  Bradbury  and 
Trager,  1967).  Peritrichous  ciliates  swarm  at  ecdysis  of  the  gammarid  amphipod 
host  (Fenchel,  1965).  The  bryozoan,  Triticclla  korcni,  metamorphoses  only  on 
the  cuticle  of  a  recent  postmolt  Calocaris  (Thalassinidea)  (Strom,  1969),  and  the 
colony  produces  embryos  just  prior  to  the  annual  molt  of  the  host  (Strom,  1969; 

121 


122  ARMAND  M.  KURIS 

Eggleston,  1971).  Total  abundance  of  epi/oic  hydroids,  bryozoans  and  barnacles 
is  greater  on  crabs  (Batliyncctcs)  in  late  postmolt  (Ci_3)  plus  intermolt  (C4)  molt 
stages  than  on  early  postmolt  (Aj-BL.)  stages  (Lewis,  1976).  The  present  study, 
employs  the  molt-staging  scheme  of  Drach  (1939;  Drach  and  Tchernigovtzeff, 
1967)  to  examine  epibiont-host  synchrony  in  detail. 

The  intricate  exoskeletal  morphology  of  an  arthropod  offers  unique  opportuni- 
ties for  studies  of  habitat  preference  and  utilization.  Here  is  a  habitat  which  is 
varied,  yet  standardized.  A  given  pit  or  groove  varies  in  a  manner  that  lends 
itself  to  a  reasonably  simple  quantitative  description.  Studies  of  habitat  exploita- 
tion and  selection  for  epizoic  forms  on  crustacean  hosts  have  rarely  (Walker, 
1974)  reached  the  level  of  sophistication  demonstrated  in  studies  of  water  mites  on 
aquatic  insects  (Efford,  1965;  Mitchell,  1967,  1968;  Lanciani,  1970,  1971;  Davids, 
1973). 

Although  egg  predators  of  crustaceans  are  common  (Kuris.  1971),  few  popula- 
tion studies  have  been  conducted.  Humes  (1942)  and  Hopkins  (1947,  1970) 
describe  the  life  cycle  of  Carcinonemertcs  carcinophila  on  the  blue  crab,  Callinectes 
sapidns,  in  detail.  Wickham  (1977)  describes  a  new  distinctive  species  of  Carci- 
noncmertes  from  Cancer  uiagistcr  and  indicates  (Wickham  and  Fisher,  1978)  that 
it  is  responsible  for  considerable  brood  mortality  in  this  commercially  important 
species. 

Other  than  the  original  description  from  the  kelp  crab  Pugcttia  prodncta  (Coe, 
1902)  and  a  host  (Eiiphylax  dovci)  and  range  extension  to  Payta,  Peru  (Humes, 
1942),  C.  cpialti  is  unstudied.  Carcinonemertcs  cpialti  occurs  on  //.  orcgoncnsis 
at  13  localities  from  Bahia  San  Quintin,  Baja  California,  Mexico,  to  Page's  Lagoon, 
British  Columbia,  Canada  (Kuris,  1971).  The  geographic  distribution  on  the 
Pacific  Coast  of  North  America  appears  to  be  continuous  between  these  two 
localities. 

MATERIALS  AND  METHODS 

Field  studies  were  conducted  at  Campbell  Cove,  Bodega  Harbor,  Sonoma 
County,  California,  during  1969-1971.  Additional  collections  were  made  during 
the  summers  of  1973-1975.  Material  for  some  studies  was  sometimes  collected 
elsewhere  in  Bodega  Harbor.  Hemigrapsus  orcgoncnsis  was  collected  monthly 
from  randomly  placed  removable  substrate  traps  (sampling  program  detailed  in 
Kuris,  1971),  and  the  nemertean  populations  were  censused.  Supplementary  host 
samples  were  collected  by  hand  and  at  random  without  regard  to  size,  sex  or  re- 
productive condition. 

During  April-May,  1969,  and  June-July,  1970,  a  survey  of  26  populations  of 
H.  oregoncnsis  was  conducted  along  a  transect  from  Bahia  San  Quintin,  Baja 
California,  Mexico,  to  Ucluelet,  Vancouver  Island,  British  Columbia,  Canada. 
These  collections,  of  75-150  adult  crabs  each,  greatly  extended  the  geographic  range 
of  C.  cpialti.  Station  records  are  given  in  Kuris  (1971). 

From  all  crabs  the  following  was  recorded:  carapace  width  to  0.1  mm,  taken 
with  a  vernier  caliper  at  the  notch  immediately  anterior  to  the  third  lateral  spine; 
sex;  and  molt  cycle  stage  according  to  the  scheme  of  Drach  (1939)  and  Drach  and 
Tchernigovtzeff  (1967).  Criteria  and  techniques  used  for  molt  stage  assignment 
are  given  in  Kuris  (1971).  For  ovigerous  female  crabs  the  embryogenic  process 


BIOLOGY  OF  CARCIXOXEMERTES  EPIALTI  123 

was  divided  into  20  egg  development  stages  based  on  cell  number,  amount  of  yolk 
remaining  and  appearance  of  various  embryonic  structures  (Kuris,  1971). 

All  hosts  were  sampled  by  inspection  of  tbe  external  surface  of  the  exoskeleton. 
Special  attention  was  paid  to  the  branchial  chamber,  the  sternal-abdominal  furrow 
and  the  pereiopod  axillae.  As  the  crustacean  exoskeleton  is  a  very  complex  but 
highly  standardized  structure,  site  specificity  of  C.  cpiolti  was  detailed  by  sub- 
dividing the  host's  surface  into  150  potential  sites  on  male  crabs  and  160  sites  on 
female  crabs.  Adult  nemerteans  and  their  eggs  were  only  observed  on  ovigerous 
female  crabs.  Adult  nemerteans  were  removed,  then  sized  and  sexed  using  a  com- 
pound microscope. 

Transmission  of  juvenile  nemerteans  was  tested  experimentally.  Lightly  in- 
fested hosts  were  examined  daily,  and  all  visible  nemerteans  were  removed.  These 
hosts  were  regarded  as  clean  when  no  nemerteans  were  recovered  on  three  suc- 
cessive days.  One  group  each  of  three  individual  cleaned  males,  females  with  ripe 
ovaries  (pre-ovigerous)  or  ovigerous  females  with  broods  in  an  early  stage  of 
embryogenesis  was  exposed  to  a  single  heavily  infested  (10  +  nemerteans)  male 
crab.  Thus,  four  crabs,  one  infested,  and  three  cleaned,  were  confined  together  in 
perforated  14  mm  X  10  mm  X  4  mm  hard  plastic  boxes  maintained  in  running 
sea  water.  Controls  for  each  of  these  three  combinations  were  run  simultaneouslv 

j 

by  using  a  cleaned  male  in  place  of  the  infested  male.  A  gravel  substrate  and  a  fe\v 
small  rocks  for  cover  were  provided  in  each  transmission  box.  Crabs  were  fed 
Ulva  everv  other  day.  All  crabs  were  marked  individually  by  a  tattoo  method 

"  *  j  * 

(Kuris,  1971).  The  transmission  experiments  were  conducted  in  July  and  August, 
during  the  period  of  peak  nemertean  abundance. 

The  number  of  nemerteans  on  the  infested  and  clean  crabs  were  recorded  on  the 
experimental  days  7  and  14.  The  experiments  were  terminated  on  experimental 
day  14,  at  which  time  the  crabs  were  dissected  and  exhaustively  searched  for  the 
presence  of  nemerteans.  The  entire  transmission  experiment  was  replicated  once. 

RESULTS 
Host  specificity 

In  the  Campbell  Cove,  Bodega  Harbor,  study  area,  Carcinonemertes  cpialti  is 
regularly  found  on  Hemigrapsus  orcgoncnsis,  with  H.  nudits  serving  less  frequently 
as  a  host.  In  the  lower  reaches  of  the  intertidal,  Carcinonemertes  species  also  oc- 
cur regularly  on  Cancer  antennarius,  C.  anthonyi,  and  C.  proditctns  juveniles  and 
adults. 

Important  negative  records  based  on  hundreds  of  observations  include  Pachy- 
grapsus  crassipcs,  and  the  anomurous  crab,  Petrolisthes  cinctipcs.  No  nemerteans 
were  ever  found  on  the  surface  of  45  juvenile  Pugcttia  producta  from  the  shallow 
subtidal  regions  adjacent  to  the  study  area.  A  search  of  about  50  female  ovigerous 
P.  producta  from  elsewhere  along  the  Sonoma  coast  produced  a  single  positive 
record  (recovered  by  R.  I.  Smith  and  examined  by  the  author).  However,  P. 
producta  from  the  Santa  Barbara  Channel  is  more  frequently  infested  by  C.  epialti. 

Life  cycle 

Extrusion  of  an  egg  mass  by  a  nonovigerous  female  crab  signals  the  start  of  the 
reproductive  phase  of  the  nemertean  life  cycle.  Juvenile  nemerteans  ensheathed  on 


124 


ARMAND  M.  KURIS 


Carcinonemerles  epialti    population  structure 


2.0  3.0  4.0 

Nemertean   length  (mm) 


6.0 


All 


n  -  Juveniles  on  non-ovigerous  hosts 

•  =  On  eggmass,  not  sexed 

H  =  On  eggmass  , 

S  =  On  eggmass  , 

E3  =  On  exoskeleton  ,  regressing 

FIGURE  1.  Size-frequency  histogram,  representing  the  population  structure  of  Carcinone- 
mcrtes  epialti  during  the  period  of  host  egg  development.  Average  sizes  of  sexable  males 
(short  dash  line),  females  (long  dash  line)  and  the  entire  sample  (solid  line)  are  superimposed 
on  histogram. 


BIOLOGY  OF  CARCINONEMERTES  EPIALTI  125 

regions  of  the  exoskeleton  remote  from  the  host  egg  mass  exsheath  and  migrate  to 
the  pleopods  or  sternal  surface  of  the  thorax  and  abdomen.  Here  each  individual 
constructs  and  inhabits  a  mucous  tube.  Occasionally,  both  sexes  may  be  found 
within  the  larger  female  tube.  A  gravid  C.  cpialti  female  deposits  her  eggs  in  the 
posterior  portion  of  her  mucous  tube.  No  nemertean  egg  tubes  are  found  on  crabs 
with  embryogenesis  less  advanced  than  initiation  of  thoracic  limb  development 
which  is  reached  18  days  after  host  oviposition  (Kuris,  1971).  Embryogenesis 
advances  until  thoracic  limb  buds  are  large  before  hatching  nemertean  eggs  are  ob- 
served. At  10-12°  C  thoracic  limb  bud  development  takes  8  days  (Kuris,  1971). 
This  interval  estimates  the  duration  of  nemertean  embryogenesis.  Nemertean  ovi- 
position may  proceed  for  the  next  25  days,  until  host  eclosion.  Nemertean  egg 
hatching  may  continue  five  days  subsequent  to  the  hatching  of  the  crab  eggs. 

Growth  of  the  nemerteans,  on  the  host's  egg  mass  from  the  time  of  host  egg 
deposition,  is  rapid  (Fig.  1).  From  the  average  juvenile  length  of  1.0  mm,  female 
worms  grow  to  an  average  adult  size  of  3.9  mm  24-30  days  after  the  host  becomes 
ovigerous.  Males  grow  more  slowly,  reaching  an  average  of  2.5  mm  18-24  days 
after  egg  extrusion.  Thirty  days  from  the  time  of  egg  deposition,  some  of  the 
nemerteans  begin  to  regress,  even  to  leave  the  egg  mass  and  retire  to  the  other 
sites  on  the  crab.  By  the  time  the  zoeae  hatch,  the  average  sexable  adult  females 
are  2.8  mm;  the  males,  1.6  mm.  Both  of  these  values  probably  over-estimate  the 
size  of  adult  worms,  as  regression  below  1.5  mm  makes  sexing  difficult.  The 
average  size  of  all  worms  on  egg-bearing  females  reaches  a  maximum  of  2.5  mm  by 
30  days  after  egg  deposition,  and  then  declines  to  1 .4  mm  by  the  time  the  host's 
brood  hatches,  44  days  after  having  become  ovigerous. 

The  nemertean's  modified  (Hyman,  1951)  pilidium  larva  appears  to  be  plank- 
tonic  for  an  unknown  period  of  time.  Ultimately,  this  dispersal  stage  must  settle 
on  a  crab  host  and  transform  to  a  juvenile. 

Crab  hosts  of  either  sex,  mature  or  not,  may  become  infested  with  juvenile  C. 
epialti.  However,  only  when  ovigerous  or  pre-ovigerous  crabs  are  infested  may  the 
life  cycle  be  completed. 

Some  of  these  larval  and  juvenile  nemerteans  reach  pre-ovigerous  female  hosts. 
The  sites  occupied  by  juveniles  on  such  crabs  are  similar  to  those  occupied  by 
juveniles  on  nonovigerous  hosts.  However,  a  day  or  two  after  a  pre-ovigerous  host 
undergoes  oviposition,  virtually  all  the  juvenile  nemerteans  ensheath  and  migrate 
to  the  vicinity  of  the  host's  egg  mass.  Here  the  nemerteans  begin  a  period  of  rapid 
growth,  sexual  differentiation  and  maturation.  Copulation  presumably  occurs  when 
the  male  nemertean  enters  the  female's  mucous  sheath. 

As  the  host's  eggs  near  the  date  of  hatching,  the  nemerteans  cease  to  grow 
(Fig.  1).  Some  worms  leave  the  egg  mass,  frequently  migrating  to  sites  within 
the  branchial  chamber  of  the  host.  The  anterodorsal  surface  of  the  host's  branchial 
chamber  is  frequently  occupied  by  these  worms.  Here  they  ensheath,  decrease  in 
size,  and  become  reproductively  inactive.  Soon  they  are  indistinguishable  from 
juvenile  nemerteans.  Some  of  the  post-reproductive  worms  may  die  rather  than 
regress.  Large,  seemingly  moribund  individuals  are  seen  shortly  after  eclosion  of 
the  host  brood.  These  worms  may  merely  be  undergoing  negative  growth,  how- 
ever. Whether  secondarily  reduced,  post-reproductive  worms  are  capable  of 
another  reproductive  period  is  unknown. 


126 


ARMAND  M.  KURiS 


TAHLK  I 

Experimental  transfer  of  juvenile  ('.  epia.lti.from  heavily  infested  male  c-  ib.i  (donors) 
to  uninfested  male  and  female  (ovigerous  and  nonovigerous)  hosts. 


Experimental 
combination 

Number 
of  hosts 

Number  of 
worms 
removed 
prior  to 

start 

Number  of  worms 
observed  on  day 

Number  of 
worms 
after 
dissection 

Number 
of  worms 
unaccounted 
for 

Mean  worm 
density  on 
recipients 
day  14 

0 

7 

14 

Donors 

2 



54 

31 

14 

22 

18 



Recipient  males 

6 

3 

0 

7* 

11 

14 

•  — 

2.3 

Donor  control 

1 

0 

0 

0 

0 

0 

0 

— 

Recipient  male 

3 

6 

0 

1 

1 

1 

— 

0.3 

controls 

Donors 

2 

— 

100 

75 

57 

69 

19 



Recipient  non- 

6 

4 

0 

1 

6 

12 

— 

2.0 

ovigerous 

females 

Donor  control 

1 

0 

0 

0 

0 

0 

0 

0 

Recipient  non- 

3 

5 

0 

3 

3 

3 

— 

1.0 

ovigerous 

female  controls 

Donors 

2 

0 

115 

105 

68 

75 

12 



Recipient 

6 

2 

0 

5 

12** 

14 

— 

4.0 

ovigerous 

females 

Donors  control 

1 

0 

0 

0 

0 

0 

0 

— 

Recipient 

3 

3 

0 

2 

2* 

2 

— 

1.0 

ovigerous 

female  controls 

*  One  recipient  died  prior  to  observation  date,  data  for  day  14  based  on  2  donors  and  5  re- 
cipients. 

**  One  donor  died  prior  to  observation  date,  data  for  day  14  based  on  1  donor  and  3  recipient 
crabs. 

Based  upon  microscopic  examination  of  the  gut,  host  eggs  appear  to  be  the  only 
source  of  nutrition  for  the  adult  worms.  Juvenile  worms  appear  to  have  empty 
digestive  tracts.  Thus,  it  is  not  surprising  to  find  that  growth  of  juvenile  worms 
on  male  hosts,  juvenile  females,  or  nonovigerotis  adult  females  is  quite  restricted. 
The  size  range  of  juvenile  nemerteans  from  these  sources  is  only  0.4  to  1.6  mm. 
These  juveniles  are  considered  to  he  essentially  phoretic  on  the  host  crab.  Since 
the  newly-hatched  larvae  are  0.2  mm  long,  little  food  seems  to  be  required  for 
transformation  to  the  juvenile  form,  and  subsequent  maintenance  on  the  host  exo- 
skeleton.  The  source  of  energy  for  this  maintenance  remains  unknown. 

Transmission 

It  is  likely  that  there  are  two  modes  of  transmission  for  C.  cpialti.  The  free- 
swimming  larval  stage  facilitates  interhost  transmission.  Incidents  of  direct  con- 
tact between  hosts,  followed  by  transferral  of  juvenile  worms,  may  also  enable 
transmission  to  occur.  Direct  transmission  was  tested  experimentally. 


BIOLOGY  OF  CARCINONEMERTES  EI'LILTI 


127 


Carcinonemertes  epiolti  on   Hemigrapsus  oregonensis 
Seasonal  Changes 


88     55 

?  ?  n.ov.  74     51 
SJov.        11        9 


29     57      31  46 

24     53     40  45 

-        2         7  32 


13 
18 
1  1 


30 
20 

1 


lOOr 


a> 

-» — 

1/1 
OJ 


c 
a> 
o 
\_ 
a> 
a. 


20  - 


40 

c 
a> 
-o 

=    30 


C 
O 
QJ 


20 


10 


0 


A 

/\ 


Both  sexes 
d'cf 


Ovigerous   ?? 
Non-ovigerous  ? 


Jun   Jul       Aug       Sep 
1969 


Oct        Nov        Dec      Jan       Feb        Mar      Apr       May      Jun 
1970 


FIGURE  2.  Seasonal  pattern  of  the  percentage  of  infestation  (top)  and  mean  burden  (mean 
density  per  host)  (bottom)  for  all  hosts,  ovigerous  females,  nonovigerous  females  and  males 
over  8  mm,  at  Bodega  Harbor. 


128  ARMAND  M.  KURIS 

Table  I  shows  that  significant  transfer  ((.i-test,  day  14,  P  <  0.05)  of  juvenile 
nemerteans  may  occur  between  hosts  and  suggests  that  ovigerous  hosts  may  elicit 
more  transferrals  than  nonovigerous  hosts.  The  occasional  nemerteans  seen  on  the 
externally  cleaned  and  unexposed  control  crabs  are  probably  derived  from  sites 
hidden  within  the  host's  branchial  chambers,  inaccessible  to  the  removal-trapping 
method  of  cleaning  crabs. 

Infestation  frequency  and  nemertean  density 

Seasonal  variation.  For  741  crabs  greater  than  8.0  mm  wide,  the  overall  infesta- 
tion rate  for  1969-70  at  Bodega  Harbor  was  36.3%.  Burden,  the  mean  density 
per  host  (including  uninfested  hosts),  was  3.96. 

Through  the  sample  year,  the  infestation  level  of  C.  cpialti  on  H.  oregonensis 
varied  from  28%  to  57%.  Figure  2  shows  that  the  overall  infestation  rate  actually 
remained  steady  at  30-40%,  except  for  the  September  and  October  samples,  which 
reached  57%  and  46%,  respectively.  This  rise  occurs  when  the  host  population 
is  reproductively  inactive  while  undergoing  the  final  ecdysis  prior  to  the  onset  of 
the  winter  anecdysial  period  (Kuris,  1971).  The  maturation  of  large  juvenile  and 
prepuberty  female  hosts  also  occurs  at  this  time ;  so  virtually  all  the  crabs  over  8.0 
mm  are  adults  in  the  coming  winter  reproductive  season.  The  nemertean  density 
per  host  reflects  this  pattern,  rising  to  an  October  peak  density  of  11  worms  per 
host.  Excepting  the  autumn  samples,  worm  burden  ranges  from  0.5  to  4.0  worms 
per  host.  Surprisingly,  the  peak  period  of  crab  reproduction,  November  to 
February  (Kuris,  1971),  is  the  interval  of  lowest  nemertean  density  (Fig.  2). 

Host  reproduction.  The  importance  of  host  reproduction  to  population  dynamics 
of  C.  epialti  is  seen  in  Figures  2  and  4.  In  all  seasons  the  frequency  of  ovigerous 
female  crabs  harboring  nemerteans  is  higher  than  that  of  nonovigerous  females  or 
males.  With  the  exception  of  the  January  sample,  this  is  also  true  of  the  average 
nemertean  density  per  host  crab  over  8  mm. 

As  host  eggs  proceed  through  embryogenesis,  an  increase  in  nemertean  preva- 
lence and  density  might  be  expected  on  ovigerous  crabs  (Table  II).  The  percent- 
age of  infestation  increases  slightly,  from  62.9%  of  broods  in  embryogenic  stages 
to  73.7%  of  broods  in  late  stages,  with  slight  fluctuations  at  intermediate  stages. 

Burden  (b)  remains  essentially  the  same  (4.2-4.4)  from  early  through  late  middle 

egg  development  stages.  However,  b  then  rises  sharply,  to  8.11  in  late  stage 
broods. 

During  host  embryogenesis  the  wrorms  anchor  their  mucous  sheaths  and  feed 
while  protruding  from  the  open  end.  The  nemerteans  are  able  to  feed  on  the  eggs 
if  their  sheath  is  entwined  among  the  host  egg  mass  or  is  attached  to  the  abdominal 
appendages  or  the  sternal  surfaces  of  the  abdomen  and  the  thorax.  On  nono- 
vigerous adult  female  crabs  only  73.3%  of  the  nemerteans  are  found  in  the  vicinity 
of  the  egg  mass.  However,  almost  immediately  after  deposition  of  the  host's 
brood,  95.9%  are  near  the  egg  mass.  This  distribution  pattern  remains  almost 
constant  for  the  first  26  days  of  host  embryogenesis.  Towards  the  end  of  the 
brooding  period,  a  gradual  withdrawal  to  other  sites  is  evident  (Table  II).  Only 
64.7%  of  the  nemerteans  remain  at  sternal  locations  on  post-ovigerous  female  crabs. 

Host  molt  cycle.    Figure  3  shows  the  changes  in  nemertean  burden  on  similar- 


BIOLOGY  OF  CARCIXONEMERTES  EPIALTI 


129 


60- 


20. Ot 


16.0-19.9 
x 


D,. 


3-4 


H .  oregonens/s     molt    stage 

FIGURE  3.  Average  burden  (mean  density  per  host)  of  Carcinoncmertcs  epialti  on  different 
sized  male  Hcmigrapsus  orcgoncnsis  in  different  stages  of  the  molt  cycle. 

sized  male  hosts,  with  successive  molt  stages.  The  average  burden  is  seen  to  rise 
sharply  from  Drach  molt  stages  A  to  C3  or  C4.  However,  from  C4  or  D0  to  D! 
there  is  a  sharp  drop  in  average  density  per  host ;  this  is  followed  by  an  equally 
sharp  rise  in  late  premolt,  Do-D4. 

Host  size.  Figure  4  shows  that  both  nemertean  incidence  of  infestation,  and 
the  average  burden,  increase  dramatically  with  increasing  host  size.  Ovigerous 
females,  while  showing  some  size  effects  (Fig.  4  bottom),  do  not  show  as  sharp  an 
increase  in  average  density  with  increasing  size  as  do  males  and  nonovigerous 
females.  The  incidence  of  infestation  among  different  size  classes  is  significant 

TABLE  II 

Percentage  infestation  (c/(i),  average  burden  (5),  and  site  preferences  of  C.  epialti  on  origerous 
crabs  through  the  course  of  embryogenesis.     Post-ovigerous  crabs  are  also  included. 


Egg  development  stages 
(grouped) 

n 

%i 

b 

Percentage  of  nemerteans 
on  egg  mass  and 
on  thoracic  and 
abdominal  sterna 

Percentage  of 
nemerteans  at 
other  sites 

Early  (1  to  12  days) 

35 

62.9 

4.2 

95.9 

4.1 

Early  middle  (13  to  26  days) 

58 

74.1 

4.4 

95.7 

4.3 

Late  middle  (27  to  37  days) 

38 

76.3 

4.3 

88.1 

11.9 

Late   (38  to  43  days) 

10 

73.7 

8.1 

73.2 

26.8 

Post-ovigerous 

1  1 

72.7 

17.0 

64.7 

35.3 

(duration  uncertain) 

130 


ARMAND  M.  KUKIS 


C.  epialti   on     d7     Crabs 


Hosts  8.0-ll.9mm 


N  %  b 

212          8.5          0.15 


40i 

101        Hosts  12.0-15.9  mm 


0 
5- 

0 

5H 

0 


82         58.5          4.66 


Hosts  16. 0-19. 9mm 

^  i  j  « «  •  • •  i 

Hosts  20.0  mm 


40         90.0        20.40 


17       100.0        43.76 

//   •    // //  •    //  • 


o> 
> 

i_ 
CD 

(S> 

.a 
o 

(/> 

O) 

I/) 
o 
o 


cu 

JD 

e 

Z5 


10       20       30       40       50       60       70       80      90       !24        I35       I96    2I2 

C. epialti  on  Non-ovigerous     ^    Crabs 


I80i 

20- 
iO  I        Hosts  8.0 -1 1. 9  mm 


214         15.9          0.24 


Hosts  I2.0-I5.9mm 


80        51.2 


3.38 


Hosts  16.  0-19.  9  mm 


20         90.0        14.30 


Hosts  20.0  mm 


3       100.0        23.67 


10       20       30       40       50       60      70       80      90 


30 


C.  epialti  on  Ovigerous 


Hosts  8.  0-11.9  mm 


Hosts  12.  0-15.  9mm 


5 

0 


,k^! 

Hosts  16. 
liUbi       i        j 


.0  -19.9  mm 


Crabs 

37  51.4  2.41 

23  78.3  2.30 

13  84.0  11.0 


10        20       30       40       50       60      70       80      90 

Number   of   nemerteans  on   host 


BIOLOGY  OF  CARCINONEMRRTES  EP1ALTI  131 

(G-test,  P  <  0.001)  for  males  and  noimvigcnms  females  and  for  ovigerous  females 
(P  <0.05). 

Within  each  size  class  there  is  no  significant  difference  hetween  the  three 
reproductive  classes  except  for  the  8.0-11.9  mm  size  class  (P  <  0.005).  Relatively 
high  levels  of  infestation  among  small  (8.0-11.9  mm)  ovigerous  females  account 
for  both  the  generally  high  infestation  rate  of  ovigerous  females  compared  with 
males  and  nonovigerous  females,  and  the  significant  difference  in  incidence  seen 
among  the  reproductive  categories  for  the  smallest  size  class. 

Two  changes  in  the  pattern  of  nemertean  abundance  occur  as  host  size  in- 
creases. The  frequency  of  uninfected  crabs  drops  sharply  (91.5%  to  0.07c)  with 
increasing  size.  Also,  the  frequency  of  heavy  infestations,  b  >  9,  increases  strongly, 
from  0.2%  to  65.0%.  These  trends  with  increasing  size  result  in  contagious  dis- 
tribution patterns;  the  variance  to  mean  ratio  (coefficient  of  dispersion)  greatly 
exceeds  one  in  all  groups  over  12  mm.  Even  in  the  8-12  mm  size  classes,  the  co- 
efficient of  dispersion  is  over  one,  indicating  that  these  samples  are  also  clumped. 
Tests  for  goodness  of  fit  (x2,  Sokal  and  Rohlf,  1969)  result  in  highly  significant 
differences  from  the  expected  Poisson  distribution  in  all  4  mm  size  classes  for  male, 
ovigerous  female  and  nonovigerous  female  hosts. 

Site  specificity 

The  distribution  of  nemerteans  is  analyzed  here  for  male  crabs  only.  Nemer- 
teans  on  nonovigerous  crabs  have  a  similar  pattern  of  occurrence,  differing  only  in 
the  details  of  female  versus  male  sternal  anatomy. 

On  male  crabs  fifty  of  the  sixty  investigated  sites  harbored  C.  epialti  with  some 
regularity.  The  most  frequented  sites  include  the  anterior  face  of  the  arthrodial 
membrane  at  the  base  of  the  coxa  of  the  fourth  walking  leg,  the  posterior  face  of 
the  equivalent  membrane  of  the  cheliped.  and  the  ventral  angles  of  the  axillae  be- 
tween the  second  and  third,  and  third  and  fourth  walking  legs.  Other  points  at  the 
bases  of  the  limbs  were  only  slightly  less  commonly  utilized  as  sites.  In  general, 
locations  in  the  second  axilla,  between  the  first  and  second  walking  legs,  were  the 
least  commonly  inhabited  sites  on  the  limb  bases.  In  the  sternal-abdominal  furrow, 
the  anterior  sternal  sutures  of  the  thorax,  the  bases  of  the  copulatory  pleopods,  and 
the  anterior  segments  of  the  abdomen  were  often  frequented.  In  the  gill  chamber, 
the  fourth  and  fifth  thoracic  epimera  and  the  vicinity  of  the  pericardial  sacs  were 
commonly  utilized  sites. 

A  single  case  of  internal  infection  of  H.  oregonensis  by  C.  epialti  was  observed. 
Three  juvenile  nemerteans  were  found  in  the  posterior  portion  of  the  host's  in- 
testine. Occasionally,  juvenile  nemerteans  are  wedged  deeply  into  the  apodemes 
originating  from  the  branchial  region  of  the  thorax.  Superficially,  these  resemble 
internal  infections.  Exsheathed  juveniles  are  occasionally  found  actively  wandering 
about  the  host  surface. 

FIGURE  4.  Size  frequency  histogram  for  Carcinonemertes  cpialti  on  different  size  classes  of 
351  male,  317  nonovigerous  and  73  ovigerous  female  Hcmiyrapsus  oregonensis.  N.  is  the 
sample  size;  %i  is  the  percentage  infested;  and  b,  mean  burden,  is  mean  density  per  crab 
(including  uninfected  crabs).  For  all  male  crabs  %i  =  36.0%,  b  =  5.70;  for  all  nonovigerous 
females  %\  =30.3%,  b  =  2.14;  for  all  ovigerous  females  %'\  =  65.8%,  b  =3.90. 


132 


ARMAND  M.  KURIS 


TABLE  III 

Site  utilization  hy  (".  epiulti  on  mule  II.  nrejjoimisis  at  different  host  sizes.  All  densities  of  worm 
infestation  are  included.  See  text  for  descriptions  of  site  regions  I- IV.  The  percentage  of  nemer- 
tetiHS  in  a  region  is  in  parentheses;  b  is  the  number  of  nemerteans;  i  is  the  number  of  hosts  infested. 
GH  =  93.268,  P  <  0.005;  an  a  posteriori  .S'77'  shows  that  the  two  smallest  and  three  largest  size 
classes  constitute  homogeneous  sets. 


b  at  sites 

Host  size 

Sh 

2b 

(in  mm) 

I 

II 

III 

IV 

i 

8-11.9 

4  (5.6) 

3     (4.2) 

64  (88.9) 

1  (1.3) 

72 

14 

1.64 

12-15.9 

28  (4.1) 

80  (11.6) 

566  (82.1) 

15  (2.2) 

689 

83 

8.30 

16-19.9 

136  (7.5) 

418  (22.9) 

1243  (68.1) 

27  (1.5) 

1824 

68 

26.82 

20  + 

182  (7.5) 

578  (24.0) 

1614  (66.9) 

39  (1.6) 

2413 

76 

31.75 

350 

1079 

3487 

82 

4998 

271 

18.44 

To  examine  the  relationship  between  host  size  and  site  specificity,  potential  sites 
were  grouped  into  four  regions.  (I)  branchial  chamber-pericardia! ;  (II)  sternal- 
abdominal  furrow;  (III)  interlimb  axillae;  (IV)  miscellaneous  (mouthparts,  ex- 
posed body  surface).  Table  III  indicates  that  only  region  II  shows  a  progressive 
increase  in  the  percentage  of  site  utilization  of  C.  cpialti  with  increasing  host  size. 
The  heterogeneity  G-test  statistic  (Gn;  Sokal  and  Rohlf,  1969)  is  highly  significant; 
size-specific  site  utilization  is  not  homogeneous.  As  there  is  significant  hetero- 
geneity among  size  classes,  an  a  posteriori  test  by  a  simultaneous  test  procedure 
(STP)  of  the  different  sizes  for  goodness  of  fit  (Sokal  and  Rohlf,  1969)  is  used  to 
locate  the  source  of  the  heterogeneity.  The  results  of  the  STP  using  the  G-statistic 
indicate  that  a  highly  significant  difference  in  the  frequency  of  the  nemertean  at 
certain  sites  occurs  between  small  (<  15.9  mm)  and  large  (>  16.0  mm)  crabs. 
Apparently,  some  of  the  sites  in  the  sternal-abdominal  furrow  are  not  available  to 
the  nemerteans  on  small  crabs.  Presumably  due  to  spatial  considerations,  these 
sites  become  available  on  crabs  over  16.0  mm. 

Changes  in  site  utilization  in  relation  to  nemertean  density  were  analyzed  in  a 
similar  fashion  to  the  site-host  size  relationship.  Infested  crabs  were  apportioned 


TABLE  IV 

Shift  in  site  preference  with  changes  in  nemertean  density  on  H.  orcgonensis.  All  sizes  of  infested 
hosts  included.  Site  regions  I— IV  are  described  in  text.  The  percentage  of  nemerteans  in  a  region 
is  in  parenthesis;  b  is  the  total  number  of  nemerteans;  i  is  the  number  of  hosts  infested.  GH  =  110.186, 
P  <  0.005;  an  a  posteriori  STP  disclosed  these  homogeneous  sets:  a)  1-10,  11-19,  20-49,  b)  20-49, 
100 +  ,  c)  1-10,  11-19,  50-99. 


Range  of 

2b 

Mean 

worm 

I 

II 

III 

IV 

2b 

i 

host 

burdens 

l 

size 

1-10 

47  (8.2) 

90  (15.6) 

435  (75.5) 

5  (0.0) 

577 

156 

3.70 

15.6 

11-19 

19  (5.0) 

67  (17.7) 

286  (75.7) 

6  (1.6) 

378 

29 

13.03 

18.0 

20-49 

88  (8.5) 

217  (20.9) 

720  (69.2) 

15  (1.4) 

1040 

31 

33.55 

19.8 

50-99 

57  (5.0) 

180  (15.8) 

876  (77.0) 

24  (2.1) 

1137 

15 

75.80 

18.8 

100  + 

139  (7.5) 

525  (28.2) 

1163  (62.6) 

32  (1.7) 

1859 

12 

154.92 

20.0 

BIOLOGY  OF  CARCIXOXEMERTES  EPIALTI  133 

among  five  nemertean  density  classes  (Table  IV)  without  regard  to  host  size;  GH 
is  highly  significant.  An  a  posteriori  STP  was  performed  to  locate  the  source  of 
the  heterogeneity.  The  STP  shows  that  crabs  having  20-40  and  100+  nemerteans 
(Table  IY,  homogeneous  set  b)  have  significantly  more  nemerteans  on  region  II  and 
fewer  on  region  III  than  occur  at  other  worm  densities  (homogeneous  set  c). 
Homogeneous  set  a  shows  that  there  is  some  overlap  between  sets  b  and  c  at  these 
sample  sizes. 

DISCUSSION 

The  common  occurrence  of  Carcinoncincrtcs  epialti  on  Hemigrapsus  oregonensis 
contrasts  with  its  occasional  presence  on  H.  mid  us,  its  scarcity  on  Pugettia  producta, 
and  its  absence  on  Pachygrapsns  crassipes.  The  infrequent  infestation  of  Pugettia 
producta  (=  Epialtus  prodiictus]  by  C.  epialti  is  of  interest,  as  this  is  the  type  host 
(Coe,  1902).  At  least  along  the  Sonoma  coast,  the  specific  name  epialti  is  an  un- 
fortunate choice.  All  the  ovigerous  specimens  of  P.  producta  from  this  region  are 
infested  with  several  hundred  turbellarian  egg  predators  of  an  undescribed  species 
[Monocelisf  (Sakaji,  personal  communication)].  The  inadequately  documented 
record  of  C.  epialti  on  P.  producta  (Boolootian,  Giese,  Farmanfarmaian  and  Tucker, 
1959)  is  probably  a  misidentification  of  the  undescribed  turbellarian.  The  size  (1- 
2  mm),  and  the  activity,  "gliding  continually"  (p.  219),  both  fit  the  turbellarian, 
and  decidedly  not  C.  epialti. 

The  small  worms,  found  on  Cancer  inagister  by  MacKay  (1942)  and  probably 
misidentified  as  leeches  (Sindermann  and  Rosenfield,  1967),  are  most  likely  the 
Carcinonemertes  species  to  be  described  by  \Yickham  (1977).  Specific  identifica- 
tion of  the  Carcinonemertes  found  on  other  Cancer  species  awaits  further  study 
("YYickham,  personal  communication). 

Humes  (1942)  observed  that  20  of  the  26  host  records  for  Carcinonemertes  spp. 
available  to  him  were  for  portunid  crabs  (including  the  Peruvian  portunid  Euphy- 
la.v  dovei  as  a  host  for  C.  epialti }.  He  considered  the  littoral  Portunidae  to  be  the 
principal  hosts  for  these  worms  due  to  their  habits,  abundance,  and  habitat  prefer- 
ences. The  host  specificity  records  for  Carcinoncincrtcs  on  the  Pacific  coast  of 
North  America,  suggest  that  neither  portunids,  nor  the  behavioral  and  habitat 
characteristics  typically  associated  with  the  swimming  crabs,  are  necessarily  as- 
sociated with  nemertean  infestations. 

The  Carcinonemertidae  are  considered  to  exhibit  little  host  specificity  (Humes, 
1942).  However,  the  negative  records  on  Pachygrapsus  crassipes,  despite  the 
habitat  overlap  of  these  crabs  with  heavily  infected  host  species,  suggests  that  host 
specificity  does  play  a  part  in  governing  the  distribution  of  C.  epialti. 

A  comparison  of  the  life  cycle  of  C.  epialti  with  C.  carcinophila  (Humes,  1942; 
Hopkins,  1947)  shows  some  important  differences.  The  juvenile  stage  of  C. 
carcinophila  is  found  almost  exclusively  on  the  gills  of  the  nonovigerous  host.  This 
site  may  indicate  a  decreased  opportunity  for  interhost  transfer  of  juvenile  worms 
during  casual  contact.  However,  frequent  transfer  during  the  mating  act  seems 
feasible  since  copulation,  followed  by  a  post-mating  embrace,  is  a  lengthy  process  in 
portunid  crabs  (Hartnoll,  1969).  principal  hosts  of  C.  carcinophila. 

The  fate  of  the  post-reproductive  adult  nemertean  is  another  potentially  distinc- 
tive species  difference.  In  C.  carcinopliila  the  adult  worms  retire  to  the  gill  chamber 


134  ARMAND  M.  KURIS 

of  the  host  upon  hatching  of  the  crab's  brood  (Humes,  1942;  Hopkins,  1947,  1970). 
Here  they  can  be  distinguished  from  pre-reproductive  juveniles  by  their  bright  red 
color.  The  principal  western  Atlantic  host,  Callincctes  sapidns,  ceases  to  molt  upon 
reaching  adulthood  (Van  Engle,  1958),  and  thus  the  nemerteans  are  never  shed 
after  the  host's  eggs  hatch.  Hopkins  (1970)  feels  that  they  also  return  to  the  host's 
egg-mass  during  the  next  ovigerous  period.  Since  Hopkins  (1947)  describes  the 
post-reproductive  worms  in  the  gill  chambers  as  "large,"  there  is  an  indication  that 
the  post-reproductive  worms  do  not  regress  to  the  size  of  the  juvenile  worms  upon 
their  return  to  the  gills. 

Both  juvenile  transfer  and  larval  settlement  are  regarded  as  important  factors 
of  a  transmission  model  accounting  for  the  distribution  of  nemerteans  on  host 
crabs.  It  is  proposed  that  larval  settlement  of  a  short-lived  larval  phase  accounts 
for  the  occasional  occurrence  of  heavily  infested  crabs.  Since  Heinlgrapsus  orc- 
gonensis  spends  long  periods  of  time  aggregated  under  covering  rocks  (Kuris, 
unpublished  mark  recapture  study),  they  may  occasionally  encounter  a  dense  larval 
swarm  in  the  small  volumes  of  water  with  restricted  circulation  in  this  confined 
habitat.  The  considerably  greater  surface  area  of  larger  crabs  available  for  settle- 
ment, as  well  as  the  occurrence  of  additional  suitable  sites  for  juvenile  worms  on 
such  crabs,  may  result  in  most  of  the  heaviest  infestations  being  found  on  these 
crabs.  As  most  large  crabs  are  males,  this  would  also  account  for  the  more  frequent 
occurrence  of  large  numbers  of  juvenile  nemerteans  on  males. 

Transferral  of  juvenile  nemerteans  from  infested  to  uninfested  hosts  through 
accidental  and  mating  contact  may  be  the  means  by  which  most  crabs  with  low 
nemertean  burdens  (1  to  10)  become  infested,  as  such  contacts  are  of  short  dura- 
tion (Knudsen,  1964;  Kuris,  1971).  Also,  most  juvenile  nemerteans  are  usually 
surrounded  by  a  thin  mucous  sheath  ;  for  host  transfer  to  occur  they  must  escape 
the  sheath.  Thus,  only  a  small  percentage  of  the  population  of  nemerteans  are 
available  for  contact  transmission  at  a  given  instant. 

The  laboratory  transmission  experiments  indicate  that  juvenile  transferral  oc- 
curs between  donor  males  and  recipient  males,  nonovigerous  females,  and  ovigerous 
females.  Perhaps  such  transfers  also  occur  with  donor  nonovigerous  females. 
Howrever,  it  seems  likely  that  transfers  from  ovigerous  female  crabs  are  much  less 
likely  to  occur.  Thus,  ovigerous  females  come  to  have  a  much  higher  mean  per- 
centage of  infestation,  65.8%,  than  do  either  males,  35.6%,  or  nonovigerous  females, 
30.3%.  That  this  difference  is  a  result  of  transfer  interactions  rather  than  more 
favorable  conditions  for  larval  settlement  is  shown  by  the  higher  frequency  of  low 
nemertean  burdens  on  the  ovigerous  females  than  on  the  other  classes  of  hosts 
(Fig.  4).  Were  larval  settlement  to  be  enhanced  on  ovigerous  females,  then  the 
frequency  of  the  heavy  nemertean  burdens,  presumed  to  be  due  to  larval  trans- 
mission, would  be  greater  on  these  females. 

Most  large  crabs,  especially  those  over  16.0  mm,  are  infested  without  regard  to 
reproductive  state.  Thus,  only  small  crabs  show  the  effect  of  the  accumulation  of 
transferred  juvenile  nemerteans  as  an  increase  in  the  percentage  infestation  (Fig. 
4).  If  larval  nemerteans  do  not  settle  preferentially,  then  the  nemertean  burden  of 
egg-bearing  crabs  is  increased  over  nonovigerous  crabs  only  by  the  number  of 
nemerteans  gained  by  juvenile  transfer.  In  accordance  with  the  transmission 
model,  an  increase  in  the  number  of  nemerteans  per  ovigerous  host  over  the  period 


BIOLOGY  OF  CARCIXOXI-MERTES  EPIALTI  135 

of  host  embryogenesis  is  seen  (weakly)  in  Table  II,  since  ovigerotis  crabs  gain  but 
presumably  do  not  lose  nemerteans  through  contact  transferral. 

The  increase  in  nemertean  burden  during  postmolt  stages  is  also  in  accord  with 
the  nemertean  transmission  model.  However,  the  intermolt  decline  and  premolt 
rise  in  worm  abundance  for  all  host  size  classes  indicates  that  factors  other  than 
simple  accretion  of  nemerteans  through  time  are  operating.  Perhaps  male  and 
nonovigerous  females  lose  nemerteans  through  contact  transfer  to  ovigerous  crabs. 
Crabs  avoid  contact  during  postmolt,  and  copulation  is  perhaps  limited  to  C±-Di 
in  male  crabs  (Kuris,  1971).  Also,  selective  transmission  to  postmolt  crabs  might 
give  the  nemerteans  a  better  chance  to  locate  a  pre-reproductive  female,  or  a  pre- 
copulatory  male.  However,  preferential  settlement  on  these  stages  would  not  ac- 
count for  the  equally  dramatic  rise  in  the  abundance  of  nemerteans  in  Do-D^. 

The  strongly  host-size  dependent  distribution  of  C.  cplalti  does  not  appear  to  be 
due  to  a  nemertean  build-up  over  time  on  large  crabs  with  long  intermolt  intervals. 
Nemertean  populations  fluctuate  over  the  intermolt  period  when  crab  size  (and 
molt  cycle  duration)  is  held  constant.  More  likely,  crab  size  directly  influences 
nemertean  burdens.  Large  crabs  have  relatively  more  sites  to  offer  nemerteans,  and 
can  accommodate  larger  nemertean  populations ;  also  preferred  nemertean  sites  are 
more  spacious  on  large  crabs  and  can  support  more  worms  per  site. 

If  the  size  of  the  host  influences  the  availability  for  nemertean  habitation  of 
certain  sites  on  the  host's  exoskeleton,  then  the  percentage  of  nemerteans  on  rela- 
tively unavailable  site  should  rise  as  host  size  increases.  Such  is  the  case  (Table 
III).  However,  nemertean  density  also  influences  site  occupancy  (Table  IV). 
Crowding  at  high  density  may  result  in  some  individuals  occupying  suboptimal 
sites.  However,  those  density  classes  (2CM-0,  100+)  having  the  greatest  propor- 
tion of  worms  at  the  presumably  less  preferred  sites  of  region  II  also  have  larger 
mean  host  sizes  (Table  IV). 

Examination  of  the  interaction  between  the  effects  of  intermolt  interval,  site 
availability  and  site  preference  suggests  that  all  three  effect  the  distribution  of  C. 
epialti.  However,  the  increase  in  site  availability  with  increasing  host  size  seems  to 
be  the  most  important  factor  determining  site  occupancy. 


I  thank  Cadet  Hand,  "William  Hamner,  and  John  E.  Simmons  for  reading  my 
doctoral  thesis,  from  which  portions  of  the  present  study  are  derived.  I  am  par- 
ticularly grateful  to  Cadet  Hand,  Director  of  the  Bodega  Marine  Laboratory  of  the 
University  of  California,  for  supervising  my  thesis  and  for  placing  the  facilities  of 
the  marine  lab  at  my  disposal  for  follow-up  studies  during  the  summers  of  1973- 
1975.  I  also  thank  Dan  "\Yickham  for  valuable  discussions  and  access  to  work  in 
progress ;  Sue  Johnson  and  Pat  Lewis  for  manuscript  preparation ;  Emily  Read  for 
the  figures;  an  NIH  predoctoral  fellowship,  the  Zoology  Department  of  University 
of  California,  Berkeley  and  a  LTniversity  of  California,  Santa  Barbara  Faculty  Re- 
search Grant  for  financial  support;  and  Bari  Karp,  Jenny  Karp  and  Choupique  for 
moral  support. 


136  ARMAND  M.  KURIS 

SUMMARY 

1.  The  geographic  range  of  Carcinonemertes  cpialti  has  been  greatly  extended. 
The  worms  are  found  from  Bahia  San  Ouintin,  Baja  California,  Mexico,  to  Page's 
Lagoon,  Vancouver  Island,  British  Columbia,  Canada. 

2.  New  host  records  for  C.  cpialti  include  PI.  orcyoncnsis,  and  H.  nudus.     It  is 
rare  on  its  type  host  Pnycttia  producta.     Specimens  of  Carcinonemertes  of  uncer- 
tain affinities  are  also  found  on  Cancer  antennarius,  C.  anthonyi  and  C.  productiis. 

3.  Carcinonemertes  cpialti  adults  are  egg  predators  on  ovigerous  hosts.    Growth, 
demography  and  abundance  are  described  in  relation  to  the  embryogenic  stage  of 
the  host  brood  at  Bodega  Harbor,  California. 

4.  Nonfeeding  juveniles  are  ensheathed  on  individuals  of  both  host  sexes  over 
8.0  mm  carapace  width. 

5.  Transmission  experiments  show  that  contact  transfer  of  juvenile  nemerteans 
from  males  to  other  hosts  may  occur. 

6.  The  percentage  of  infestation  and  mean  density  peak  in  autumn  on  77.  orc- 
goncnsis  at  Bodega  Harbor. 

7.  Ovigerous  female  hosts  are  more  frequently  infested  with  C.  cpialti,  particu- 
larly at  small  host  sizes,  than  are  male  or  nonovigerous  female  hosts  at  Bodega 
Harbor.     However,  average  worm  density  on  ovigerous  females  is  low. 

8.  Mean  density  of  C.  cpialti  rises  through  late  postmolt,  declines  during  inter- 
molt  and  rebuilds  to  a  high  level  in  late  premolt  //.  orcyoncnsis  from  Bodega  Har- 
bor. 

9.  Large  crabs  have  a  higher  percentage  of  infestations  and  mean  densities  per 
infection  than   do   small   crabs.      Nemerteans   are   more   frequently   found   in   the 
sternal-abdominal  furrow  and  less  frequently  in  the  limb  axillae  on  large  crabs. 

10.  A  model  of  C.  cpialti  transmission  and  site  occupancy  is  proposed,  incor- 
porating the  influence  of  host  size,   sex,  reproductive  state,  embryogenesis,   molt 
cycle  stage  and  molt  cycle  duration  of  77.  oreyonensis  at  Bodega  Harbor.     Site 
availability  increases  with  host  size.     At  higher  densities  the  juvenile  nemerteans 
increasingly  occupy  less  preferred  sites.    Transferral  of  juvenile  nemerteans  occurs 
and   is  considered   responsible  for   the   high   frequency   of   low   infestation   levels. 
Ovigerous  females  are  more  likely  to  be  infested  but  with  low  density  infestations. 

LITERATURE  CITED 

BARNES,  H.,  AND  T.  BAGENAL,  1951.  Observations  on  Ncphrops  norvegicus  (L.)  and  on  an 
epizoic  population  of  Balanus  crcnatus  Brug.  /.  Mar.  Bin!.  Assoc.  U.K.,  30:  369-380. 

BOOLOOTIAN,  R.  A.,  A.  C.  GIESE,  A.  FARMANFARMAIAN,  AND  J.  TUCKER,  1959.  Reproductive 
cycles  of  five  west  coast  crabs.  Physiol.  Zoo}.,  32:  213-220. 

BOWERS,  R.  L.,  1968.  Observations  on  the  orientation  and  feeding  behaviour  of  barnacles  as- 
sociated with  lobsters.  /.  Exp.  Mar.  Biol.  EcoL.  2:  105-112. 

BRADBURY,  P.  C.,  AND  W.  TRACER,  1967.  Excystation  of  apostome  ciliates  in  relation  to  molt- 
ing of  their  crustacean  hosts.  II.  Effects  of  glycogen.  Biol.  Bull.,  133:  310-316. 

COE,  W.  R.,  1902.     The  nemertean  parasites  of  crabs.     Am.  Nat.,  36:  431-450. 

DAVIDS,  C.,  1973.  The  water  mite  Hydrachna  conjccta  Koenike,  1895  (Acari,  Hydrachnellae), 
bionomics  and  relation  to  species  of  Corixidae  (Hemiptera).  Ncth.  J.  Zool.,  23:  363- 
429. 

DRACH,  P.,  1939.  Mue  et  cycle  intermue  chez  les  Crustaces  Decapodes.  Ann.  Inst.  Oceanogr. 
Monaco,  19:  103-391. 

DRACH,   P.,   AND   C.   TCHERNIGOVTZEFF,    1967.      Sur    la   methode   de   determination    des   stades 


BIOLOGY  OF  CARCINONEMERTES  EPIALTI  137 

d'intermue  et  son  application  generale  aux  Crustaces.     Vie  Milieu  Scr.  A,  18 :  595-609. 
EFFORD,  I.  E.,  1965.     Ecology  of  the  walermite  Fellria  romijni  Besseling.     /.  Anim.  EcoL,  34: 

233-251. 
EGGLESTON,  D.,   1971.     Synchrony  between  moulting  in  Calocaris  inacrandrae   [Decapoda]    and 

reproduction    in    its   epibimit    Triticel/a    korcni    [Polyzoa    Ectoproctaj.    /.    Mar.    Biol. 

Assoc.  U.K.,  51 :  409-410. 
FENCHEL,  T.,  1965.     On  the  ciliate  fauna  associated  with  the  marine  species  of  the  amphipod 

genus  Gamrnarus  J.  G.  Fabricius.  Ophelia,  2 :  281-303. 

HARTNOLL,  R.  G.,  1969.     Mating  in  the  Brachyura.     Crustaceana,  16:  161-181. 
HOPKINS,  S.  H.,  1947.    The  nemertean  Carcinonemertes  as  an  indicator  of  the  spawning  history 

of  the  host,  Callinectes  sapidus.   J.  Parasitol.,  33:  146-150. 
HOPKINS,  S.  H.,  1970.    The  nemertean  Carcinonemertes  as  an  indicator  of  the  spawning  history 

of  its  host,  Callinectes  sapidus.    J.  Parasitol.,  56  Suppl.  2(1)  :  156-157. 
HUMES,   A.    G.,    1942.     The   morphology,   taxonomy,   and   bionomics   of   the   nemertean   genus 

Carcinonemertes.    III.  Biol.  Monogr.,  18:  1-105. 
HYMAN,  L.  H.,  1951.     The  Invertebrates:  Platyhelminthcs  and  Rhynchocoela.    The  Acoelomate 

Bilatcria,  Vol.  2.    McGraw-Hill,  New  York,  550  pp. 
KNUDSEN,  J.  W.,   1964.     Observations  of  the  reproductive  cycles  and  ecology  of  the  common 

Brachyura  and  crablike  Anomura  of  Puget  Sound,  Washington.    Pacific  Sci.,  18 :  3-33. 
KURIS,  A.  M.,  1971.     Population  interactions  between  a  shore  crab  and  two  symbionts.    Ph.D. 

Thesis,  University  of  California,  Berkeley,  477  pp. 
LANCIANI,    C.   A.,    1970.     Resource   partitioning   in   species   of   the   water   mite   genus   Eylais. 

Ecology,  51 :  338-342. 
LANCIANI,    C.    A.,    1971.      Host    exploitation    and    synchronous    development    in    a    water    mite 

parasite  of  the  marsh  treader  Hydrometra  myrae  (Hemiptera:  Hydrometridae).    Ann. 

Entomol.  Soc.  Am.,  64:  1254-1259. 
LEWIS,  E.  G.,  1976.     Epizoites  associated  with  Bathynectcs  snpcrbus  (Decapoda:    Portunidae). 

US.  Fish.  Wildl.  Scrv.  Fish.  Bull.,  74 :  225-227. 
MACKAY,  D.  C.  G.,   1942.     The  pacific  edible  crab,  Cancer  magistcr.    Bull.  Fish.  Res.  Board 

Can,,  62 :  1-32. 

MITCHELL,  R.,  1967.    Host  exploitation  of  two  closely  related  water  mites.  Evolution,  21 :  59-75. 
MITCHELL,  R.,  1968.     Site  selection  by  larval  water  mites  on  the  damselfly  Cercion  hierogly- 

phicum  Brown.  Ecology,  49  :  40-47. 
PAVLOVSKI,   E.   N.,    1934.     Organisms   as   the   environment.    Priroda    (Moskva')    1:   80-91    (in 

Russian). 
SINDERMANN,  C.  J.,  AND  A.  RosENFiELD,   1967.     Principal  diseases  of  commercially  important 

marine  bivalve  Mollusca  and  Crustacea.     U.S.  Fish.  Jl'ildl.  Serv.  Fish.  Bull.,  66:  335- 

385. 

SOKAL,  R.  R.,  AND  F.  J.  ROIILF,  1969.    Biometry.    Freeman,  San  Francisco,  778  pp. 
STROM,   R.,   1969.     Sexual   reproduction   in  a  stoloniferous   bryozoan,    Triticclla   koreni    (G.   O. 

Sars).    Zoo/.  Bidr.  Upps.,  38:  113-127. 
TRACER,  W.,  1957.     Excystation  of  apostome  ciliates  in  relation  to  molting  of  their  crustacean 

hosts.    Biol.  Bull.,  112:  132-136. 

VAN  ENGEL,  W.  A.,  1958.    The  blue  crab  and  its  fishery  in  Chesapeake  Bay.    Part  I.  Reproduc- 
tion, early  development,  growth  and  migration.     U.  S.  Fish.  Wildl.  Serv.  Comin.  Fish. 

Rev.,  20:  6-17. 
WALKER,   G.,   1974.     The  occurrence,  distribution  and  attachment  of  the  pedunculate  barnacle 

Octolasmis  miilleri  (Coker)  on  the  gills  of  crabs,  particularly  the  blue  crab,  Callinectes 

sapidus  Rathbun.  Biol.  Bull.,  147  :  678-689. 
WICKHAM,  D.  E.,   1977.     A  new  species  of  Carcinonemertes    (Carcinonemertidae :   Nemertea) 

with  notes  on  other  species  from  the  Pacific  Coast.    Proc.  Biol.  Soc.  IV ash.,  in  press. 
WICKHAM,  D.  E.,  AND  W.  S.  FISHER,  1978.     The  worm  Carcinonemertes  and  egg  mortality  in 

the  Dungeness  crab.  Mar.  Biol.,  in  press. 


Reference:  Dial.  Bull.,  154:  138-147.     (February,  1978) 


THE  EFFECT  OF  pH  ON  OXYGEN  CONSUMPTION  AND  ACTIVITY 
IN  THE  BATHYPELAGIC  MYSID  GNATHOPHAUSIA  IN  GENS 

T.  J.  MICKEL  i  AND  J.  J.  CHILDRESS 

Department  of  Biological  Sciences  and  Marine  Science  Institute,  University  of  California, 

Santa  Barbara,  California  93106 

The  physical  and  chemical  properties  of  the  ocean  at  a  given  depth  are  relatively 
stable ;  however,  there  are  appreciable  depth-related  gradients  of  many  parameters. 
Among  these  are  temperature,  pressure,  and  oxygen  concentration.  The  effects  of 
these  gradients  on  the  physiology  of  midwater  species  has  only  recently  begun  to  be 
investigated  (Teal  and  Carey,  1967;  Teal,  1971;  Smith  and  Teal,  1973,  Childress, 
1969,  1971,  1975;  Quetin  and  Childress,  1976).  Among  the  most  dramatic  are  the 
changes  in  oxygen  concentration  associated  with  the  oxygen  minimum  zones  that 
exist  at  intermediate  depths  in  most  of  the  world's  oceans  (Schmidt,  1925;  Sewell 
and  Fage,  1948;  Banse,  1964).  Gradients  of  pH  values  are  associated  with  oxygen 
minima,  and  pH  values  may  range  from  8.3  at  the  ocean's  surface  to  7.5  or  less  in 
the  oxygen  minimum  (more  than  a  six-fold  increase  in  acidity;  Park,  1968).  Al- 
though the  importance  of  blood  pH  is  well  known,  there  is  virtually  no  informa- 
tion available  on  the  metabolic  effects  of  water  pH.  The  crustacean,  Gnathophausia 
ing  ens,  which  resides  in  the  oxygen  minimum  layer,  was  chosen  for  the  study  of  this 
problem.  This  shrimp  is  a  lophogastrid  mysid  whose  respiratory  and  circulatory 
adaptions  to  low  oxygen  have  been  extensively  studied  (Childress,  1968,  1969, 
1971,  1975;  Belman  and  Childress,  1976).  This  report  examines  the  effect  of  pH 
on  oxygen  consumption,  oxygen  removal  from  the  respiratory  stream,  and  activity 
in  G.  ingcns. 

MATERIALS  AND  METHODS 

Specimens  of  Gnathophausia  ingens  were  collected  from  basins  off  the  coast  of 
Southern  California  at  depths  of  400  to  900  meters  with  a  midwater  trawl.  The 
animals  were  transported  to  the  laboratory  in  aerated  containers  maintained  at  ap- 
proximately 5°  C.  The  mysid  G.  ingens  was  chosen  as  an  experimental  animal, 
because  it  can  be  maintained  in  the  laboratory  for  long  periods  of  time  (Childress, 
1971).  All  of  the  experimental  animals  were  sexually  immature,  of  undetermined 
sex,  and  had  a  wet  weight  between  3  and  13  g. 

Oxygen  consumption 

Oxygen  consumption  rates  for  the  mysid  Gnathophausia  ingens  were  measured 
in  much  the  same  way  as  Childress  (1971).  Animals  were  placed  in  a  water- 
jacketed  respirometer  maintained  at  5.5°  C  and  covered  to  prevent  the  entrance 
of  light.  The  rate  of  change  of  the  partial  pressure  of  oxygen  in  the  respirometer 

1  Submitted  in  partial  fulfillment  of  requirements  for  the  degree  of  Master  of  Arts  at 
University  of  California,  Santa  Barbara. 

138 


EFFECT  OF  pit   JX   GNATHOPHAUSIA  KW 

was  continuously  monitored  with  a  ( 'lark-type-  oxygen  electrode  (Clark,  1956). 
The  electrode  was  calibrated  in  air-saturated  and  nitrogen-saturated  sea  water  (5.5° 
C)  before  and  after  each  experiment.  Any  experiment  in  which  the  nitrogen  calibra- 
tion changed  noticeably  or  the  air  calibration  changed  by  more  than  2°/c  was  not 
used. 

In  determining  the  effect  of  pH  on  respiration,  two  experimental  procedures 
were  followed.  The  first  tested  the  effect  of  pH  on  oxygen  consumption  at  low 
oxygen  partial  pressures.  It  was  also  used  to  estimate  the  limitation  of  activity  by 
oxygen  availability.  This  procedure  required  that  an  animal  placed  in  the  respirome- 
ter  at  a  specified  pH  be  allowed  to  consume  all  the  oxygen  present.  The  experi- 
ments that  followed  this  procedure  lasted  approximately  ten  hours,  depending  on 
the  size  of  the  animal.  Oxygen  electrodes  used  in  these  experiments  were  cali- 
brated in  sea  water  of  the  same  pH  as  the  experiment.  The  second  procedure  was 
designed  to  show  any  short-term  effect  of  pH  on  activity  or  rate  of  oxygen  con- 
sumption without  the  stress  of  low  oxygen.  This  procedure  involved  changing  the 
water  in  the  respirometer  at  five  hour  intervals  and  alternating  sea  water  of  pH 
7.9  with  that  of  either  pH  7.1  or  8.7.  These  pH  values  were  chosen  because  a 
tolerance  over  this  range  would  also  indicate  a  tolerance  to  fluctuations  in  pH 
which  might  occur  in  the  environment.  A  total  of  five  water  changes  was  usually 
made  during  a  single  run.  The  oxygen  electrodes  used  in  these  experiments  were 
calibrated  at  pH  7.9  and  the  effect  of  pH  on  the  calibration  later  determined.  The 
difference  in  the  calibrations  was  always  less  than  0.5^  and  was  not  subtracted  in 
calculations  of  the  respiratory  rate. 

In  order  to  maintain  constant  pH  throughout  an  experiment,  it  was  necessary 
to  buffer  the  sea  water.  Tris(  hydroxymethyl  jaminomethane  (final  concentration 
20  mg/liter)  adjusted  to  the  required  pH  with  either  HC1  or  NaOH  and  diluted 
with  distilled  water  so  as  to  be  isosmotic  with  salt  water  was  found  to  be  sufficient. 
To  determine  the  effect  of  buffering  the  sea  water,  experiments  of  approximately 
four  hours  duration  were  done  in  both  buffered  and  unbuffered  sea  water.  During 
this  period,  the  pH  of  the  unbuffered  water  did  not  change  greatly.  Oxygen  con- 
sumption rates  in  the  buffered  and  unbuffered  water  were  not  significantly  (P  > 
0.1,  n  —  8)  different. 

Bacterial  growth  was  minimized  by  the  addition  of  streptomycin  sulfate  (20 
mg/liter)  to  the  sea  water.  The  remaining  bacterial  oxygen  consumption  was  esti- 
mated by  measuring  the  rate  of  oxygen  consumption  in  the  respirometer  for  6-12 
hours  after  the  animal  was  removed.  These  rates  were  constant  and  always  less 
than  5r/(  of  the  total  measured  rate.  The  bacterial  rates  were  subtracted  in  calcu- 
lating the  oxygen  uptake  rates  of  the  animals. 

Oxygen  extracting  ability 

The  ability  of  G.  ingens  to  extract  oxygen  from  sea  water  at  different  partial 
pressures  was  calculated  from  measurements  of  the  oxygen  content  of  sea  water 
before  and  after  passing  through  the  gills.  To  measure  oxygen  in  the  exhaled 
water,  an  animal's  head  was  placed  in  a  plastic  vial  while  a  collar,  cut  from  a  rubber 
balloon,  sealed  the  animal  to  the  vial.  The  collar  was  loosely  placed  so  as  to  not 
compress  the  carapace,  and  the  flow  of  water  through  the  vial  was  checked  with  a 
nontoxic  dye.  A  microcathode  (0.0152  mm  diameter  platinum  cathode)  oxygen 


140 


T.  J.  MICKEL  AND  J.  J.  CHILDRESS 


o 

z 

LU 

o 
o 

LL! 

cr 
u_ 

UJ 


UJ 

tr 


20  - 
10  - 

0 
30 

20  - 
10 


pH    7.9 


pH      7.1 


0        20       40       60       80       100      120      140      160     180     200 

ACTIVITY   (beats  per  minute) 

FIGURE  1.  The  relative  frequencies  of  constant  activities  lasting  at  least  ten  minutes  at  pH 
7.1,  7.9,  and  8.7.  There  were  66  observations  at  pH  7.1,  206  at  pH  7.9,  and  57  at  pH  8.7. 

electrode,  chosen  because  it  is  insensitive  to  stirring,  was  placed  inside  the  vial, 
while  another  electrode  in  the  bath  recorded  the  oxygen  content  of  inhaled  water. 
A  stirrer  was  placed  in  the  bath  to  both  stir  the  electrode  and  to  keep  the  water 
evenly  mixed. 

Constant  pH  was  maintained  during  this  experiment  by  adding  Tris(hydroxy- 
methyl)aminomethane  (10  mg/liter)  to  the  water  before  the  experiment.  An 
experiment  consisted  of  placing  an  animal  in  water  of  a  specified  pH  and  reducing 
the  partial  pressure  of  oxygen  to  approximately  6-10  mm  Hg  by  bubbling  nitrogen 
through  the  water.  Constant  stirring  of  the  bath  caused  the  partial  pressure  of 
oxygen  to  slowly  return  to  approximately  100  mm  Hg  over  a  period  of  five  hours. 
The  pH  of  the  water  was  then  changed  by  slowly  adding  small  aliquots  of  either 


EFFECT  OF  PH  IN  GXATHOPIIAUSIA  141 

HC1  or  NaOH.  Each  animal  was  subjected  to  water  of  three  different  pH  values, 
the  order  being  changed  for  different  animals  to  avoid  "placement  errors".  The 
oxygen  electrodes  could  be  removed  from  the  apparatus  without  disturbing  the 
animal  and  were  calibrated  with  each  change  of  pH. 

Activity 

Activity  was  continuously  recorded  in  all  experiments.  Individuals  of  G.  ingens 
were  held  by  the  carapace  in  plexiglass  "racks"  (Quetin,  Mickel,  and  Childress, 
1978).  A  light-emitting  diode  light  source  and  miniature  photoresistor,  both  cast 
in  clear  epoxy,  were  placed  opposite  one  another  across  the  pleopods  of  an  animal. 
The  photoresistor  functioned  as  one  arm  of  a  Wheatstone  bridge.  Each  pleopod 
beat  interrupted  the  light  beam  unbalancing  the  \Yheatstone  bridge,  and  thus  gen- 
erated an  electrical  pulse.  The  pulses  were  time-averaged  by  a  cardiotachometer 
and  recorded  on  a  potentiometric  chart  recorder. 

RESULTS 
Activity  and  pH 

The  activity  of  the  mysicl  Gnathophausia  ingens  was  affected  by  pH.  The  dis- 
tributions of  constant  activities,  lasting  at  least  ten  minutes,  from  40  runs  on  eight 
individuals  is  presented  in  Figure  1.  As  shown,  the  activity  of  animals  in  water 
of  pH  7.1  and  7.9  was  remarkably  constant.  At  these  pH  values,  individuals 
either  swam  at  140  to  190  beats  per  minute  or  did  not  swim.  At  pH  8.7  individuals 
more  frequently  did  not  swim  and  were  generally  less  active  than  at  the  other  pH 
values. 

Individuals  placed  in  water  of  pH  8.7  were  observed  to  perform  extensive 
cleaning  behavior.  This  behavior  consisted  of  repeated  wiping  of  the  antennae  and 
mouthparts  with  the  first  several  pairs  of  pereiopods.  During  cleaning  activity,  the 
animal  usually  decreased  its  swimming  activity  and  this  most  likely  accounts  for 
the  trend  toward  lower  and  more  variable  activity  at  the  higher  pH. 

Pleopod  beat  was  not  affected  by  oxygen  concentration,  and  most  animals  con- 
tinued swimming  at  relatively  high  rates  for  15-30  minutes  after  the  oxygen  con- 
centration in  the  respirometer  became  immeasurably  low. 

Oxygen  consumption  and  activity 

Individuals  of  G.  ingens  were  found  to  be  capable  of  a  wide  range  of  oxygen 
consumption  rates.  Activity  of  individual  G.  ingens  had  a  profound  effect  on 
oxygen  consumption  rate  (Fig.  2).  During  a  single  run,  rates  could  vary  as  much 
as  ten-fold,  from  approximately  10  /A  Oo/(g  wet  weight -hr)  to  100  ful  Oo/(g  wet 
weight -hr).  Most  of  this  variation  could  be  attributed  to  "spontaneous"  changes 
in  the  animals'  activity.  Changes  in  activity  during  an  experiment  could  com- 
pletely mask  any  less  subtle  responses  to  the  other  parameters  being  tested.  For 
this  reason  activity  was  recorded  in  all  experiments. 

Measurements  of  changes  in  oxygen  consumption  for  short  term  changes  in 
activity  were  difficult  to  make  due  to  the  lag  in  oxygen  consumption  with  an  in- 
crease in  activity.  Oxygen  consumption  rates  were,  therefore,  calculated  for  periods 


142 


T.  J.  MICKKL  AND  J.  J.  CHI  I. DRESS 


200 


100 


Q  £     50 
I-   o» 

II 

D  +. 
CO    0> 

z  * 
o 

O    D> 


Z  — 


10 


0          0.5 

O.I 


20 


40 


60 


80 


100         120 


140 


160 


180       200 


ACTIVITY    (beats   per  minute) 


FIGURE  2.  The  relationship  between  activity  (x,  pleopod  beats/minute)  and  oxygen  con- 
sumption rate  [y,  n\  O«/(g  wet  weight -hr)|  in  Gnathophausia  iin/cns.  The  regression  line  for 
pH  7.1  is  log  y=  1.2244  + 0.0035.x,  has  an  r  of  0.737  and  is  represented  by  the  solid  line.  The 
data  points  at  pH  7.1  are  represented  by  circles.  The  regression  line  for  pH  7.9  is  log  y  = 
1 .0884  +  0.0036x,  has  an  r  value  of  0.851  and  is  represented  by  the  uniformly  dashed  line.  The 
data  points  at  pH  7.9  are  represented  by  triangles.  The  regression  line  for  pH  8.7  is  log  y  = 
1.3571  +  0.0024x,  has  an  r  value  of  0.836  and  is  represented  by  the  line  of  long  and  short  dashes. 
The  data  points  at  pH  8.7  are  represented  by  squares. 

of  constant  activity,  sustained  for  at  least  ten  minutes  and  at  partial  pressures  of 
oxygen  above  20  mm  Hg.  By  measuring  rates  at  oxygen  partial  pressures  above 
20  mm  Hg,  it  was  assured  that  tbe  animals  were  above  their  critical  partial  pressure 
of  oxygen  (Childress,  1971). 

The  relationship  of  respiratory  rate  to  activity  is  presented  in  Figure  2.  As 
shown,  the  relationship  was  found  to  be  semi-logarithmic  rather  than  linear.  This 
indicates  that  the  amount  of  oxygen  consumed  per  pleopod  beat  increases  with  the 
rate  of  pleopod  beat.  The  mean  respiratory  rate  of  nonswimming  animals,  includ- 
ing all  pH  values,  was  19.3  /xl  02/(g  wet  weight  -hr),  n  =  32,  s.d.  =  14.5.  The 
rates  for  animals  swimming  at  140  to  190  pleopod  beats  per  minute,  taken  from 
Figure  2,  are  from  39.1  to  77.5  /*!  02/(g  wet  weight -hr  ). 

O.ryf/cn  consumption  and  f>H 

The  effect  of  pH  on  respiration  in  (/'.  int/cns  was  studied  in  eight  individuals 
from  3.9  to  9.3  g  wet  weight,  in  a  series  of  40  runs.  The  large  variation  in  respira- 
tory rate  due  to  changes  in  activity  made  it  difficult  to  choose  a  representative  value 
for  the  respiratory  rate  of  an  animal  during  an  experiment.  For  this  reason,  the 
relationship  between  activity  and  respiration  for  all  animals  was  compared  at  three 
pH  values  (Fig.  2).  Respiratory  rates  for  each  animal  were  grouped  according  to 


EFFECT  OF  pH  IN  GNATHOPHAUSIA 


143 


activity  (e.g.,  all  rates  at  activity  100  to  110  pleopocl  beats  per  minute)  and  the 
means  taken.  Each  point  on  Figure  2  is  the  mean  rate  for  one  animal  at  the  mean 
activity  and  pH. 

Analysis  of  the  regression  line  of  respiration  on  activity  for  the  three  pH  values 
showed  no  significant  (P  >  0.1,  F-test)  difference  between  the  slopes  of  the  lines. 
A  test  for  homogeneity  of  variance,  however,  showed  that  the  variance  at  pH  8.7 
was  significantly  larger  (P  <  0.05,  Bartlett's  test)  from  that  at  pH  7.9  and  pH  7.1. 
Due  to  this  difference  in  variances  the  data  could  not  be  pooled,  and  therefore  a 
regression  line  for  each  pH  is  shown  (  Fig.  2 ).  The  slopes  of  all  the  regression  lines 
are  significantly  (P  <  0.001,  /-test)  different  from  zero. 

The  effect  of  pH  on  oxygen  uptake  at  oxygen  partial  pressures  of  10-30  mm 
Hg  was  studied  in  five  animals.  The  mean  respiratory  rates  and  standard  devia- 
tion in  water  of  pH  7.1,  7.9.  and  8.7  are,  respectively:  34.33  ±  15.5,  30.07  ± 
20.12,  27.16  ±  16.08  Ml  O2/(g  wet  weight -hr).  Xo  significant  (P  >  0.1,  f-test) 
difference  between  these  means  was  found. 

The  critical  partial  pressure  of  oxygen  (Pc.)  for  G.  ingens  was  found  to  be  un- 
affected by  pH  (Fig.  3).  The  Pc  values  for  11  individuals  at  three  pH  values  all 
fell  within  the  95%  confidence  interval  for  the  relationship  between  regulated  oxy- 
gen consumption  rate  and  Pc  found  for  G.  ingens  (Childress,  1971). 


0 


8 


10 


12 


14 


16 


18 


CRITICAL  PARTIAL  PRESSURE  OF  00  (mmHg) 


FIGURE  3.  The  relationship  between  oxygen  consumption  and  Pc  in  Gndthophattsia  ingens. 
The  regression  line  (x  =  4.973y  +  1.798,  solid  line),  95%  confidence  intervals  for  an  individual 
x  at  a  given  y  (dashed  lines)  and  the  solid  circles  which  these  relationships  describe  are  taken 
from  Childress  (1971).  These  observations  were  made  at  pH  values  of  approximately  8  to  8.3. 
The  data  points  from  the  present  study  are  represented  by  triangles  for  pH  7.1,  circles  for  pH 
7.9  and  squares  for  pH  8.7. 


144 


T.  J.  MICKEL  AND  J.  J.  CHILDRESS 


TABLE  I 

The  ability  of  G.  ingens  to  extract  oxygen  from  water,  at  three  pH  values,  expressed  as  the  percentage 
of  oxygen  content  of  inhaled  water.  Values  are  means  for  percentage  of  oxygen  extracted  at  oxygen 
partial  pressures  of  10—40  mm  Hg  and  40—80  mm  Hg.  The  numbers  in  parentheses  are  the  number 
of  observations  followed  by  standard  error  of  the  mean. 

pH  of  observations 


pOa  (mm  Hg) 

7.1 

7.9 

8.7 

10-40 

40-80 

10-40 

40-80 

10-40 

40-80 

Animal 

1 

62.8 

50.6 

51.7 

38.3 

58.4 

46.3 

(6,  3.6) 

(9,  0.7) 

(11,3.9) 

(10,  1.9) 

(4,  2.6) 

(13,  3.1) 

2 

70.5 

32.5 

58.7 

25.4 

58.5 

26.6 

(5,  3.6) 

(5,2.1) 

(8,  5.6) 

(10,  2.6) 

(7,  3.6) 

(18,2.1) 

3 

52.7 

54.8 

37.2 

46.0 

30.4 

48.7 

(5,  9.4) 

(12,  1.4) 

(13,  3.2) 

(19,  0.8) 

(4,  4.2) 

(21,2.1) 

4 

51.6 

42.3 

39.8 

37.8 

21.9 

36.7 

(9,  6.2) 

(13,  1.2) 

(12,3.5) 

(19,  1.5) 

(7,  2.6) 

(18,  2.3) 

5 

72.5 

63.2 

45.7 

43.8 

6.0 

52.6 

(19,  1.2) 

(17,  2.2) 

(12,2.3) 

(22,  1.8) 

(13,3.2) 

(21,  1-0) 

6 

71.9 

67.4 

53.9 

51.3 

64.3 

49.7 

(22,  1.4) 

(10,  1.8) 

(5,  4.6) 

(14,  3.0) 

(19,  1.8) 

(19,  2.3) 

Oxygen  extracting  ability  and  pH 

The  ability  of  individual  G.  ingens  to  extract  oxygen  from  water  was  measured 
in  18  runs  with  six  individuals  ranging  in  size  from  5-8  g  wet  weight.  The  re- 
sults for  the  six  animals  are  summarized  in  Table  I.  The  values  in  Table  I  are 
means  of  the  values  for  the  percentage  of  O2  extraction  in  the  ranges  of  oxygen 
partial  pressures  of  10-40  mm  Hg  and  40-80  mm  Hg.  Standard  errors  given  in 
the  table  are  not  meaningful  because,  as  can  be  seen  in  Figure  4,  the  percentage 
extraction  declines  continuously  at  higher  oxygen  partial  pressures.  Therefore, 
standard  errors  would  indicate  the  range  of  values  for  the  percentage  utilization 
over  the  oxygen  concentrations  tested,  rather  than  the  variability  of  sampled  values. 
This  would  overestimate  the  variability  found.  While  individual  shrimp  differed  in 
the  absolute  values  for  the  percentage  oxygen  extraction,  general  trends  for  re- 
sponses to  both  changes  in  pH  and  oxygen  concentration  could  be  found.  The 
results  from  a  representative  experiment  are  shown  in  Figure  4. 

The  typical  response  was  moderate  extraction  (25-55%)  at  partial  pressures  of 
oxygen  above  40  mm  Hg.  As  the  partial  pressure  of  oxygen  further  declined,  the 
percentage  extraction  increased  to  75-85%  and  in  some  cases  reached  a  peak  at 
oxygen  partial  pressures  of  10-15  mm  Hg  and  then  declined  with  a  further  decrease 
in  oxygen  concentration.  This  result  is  similar  to  that  found  by  Childress  (1971). 

No  statistical  difference  could  be  found  between  the  values  for  the  percentage  O? 
extraction  at  pH  7.9  and  8.7  (P  >  0.1),  but  each  of  the  (7.  ingens  that  was  studied 
extracted  a  significantly  (P  <  0.05,  Mann-AYhitney  test)  greater  percentage  of 
oxygen  in  water  of  pH  7.1  than  in  water  of  pH  7.9  or  pH  8.7.  As  Figure  4  shows, 
increased  Og  extraction  at  pH  7.1  occurred  over  the  wrhole  range  of  oxygen  partial 
pressures  and  apparently  was  not  related  to  the  stress  of  low  oxygen. 


EFFECT  OF  pH  IN  GNATIIOPHAVSIA 


145 


100 

• 

0 

90 

D        u 

+             0 

80 

. 

Q 

Q 

%                        a            D 

Q 
LU 

70 

a     ...     o     a            o      D           D  °                                   D 

•                D                                                                           D                                                      /-, 
_  D         ° 

1- 

Q                                                                   n     D 
H          Q 

0 

•f                                                       D                                     D 

< 

a: 

i- 

60 

-                                                        +                 j.    +                                                         D 
.            *                            °                         +     ++^+     +                                                      °     D      ° 

X 

*             •                   ++  "*"04-(-+'"''             +4. 

LU 

50 

•                                             •               •            f+.             »*x                                  ++           4 

*•        *                                                                             . 

cf 

•«                  ." 

40 

*        •    * 

* 

•    • 
\,       « 

30 

* 

20_ 

*                                 * 

j*>     i              iiiiiiiiii 

10         20         30         40         50         60         70         80 
OXYGEN   PARTIAL    PRESSURE    (mm  Hg) 


90        100 


FIGURE  4.  The  oxygen  extracting  ability  of  a  single  G.  int/cns,  at  three  pH  values,  ex- 
pressed as  the  percentage  of  oxygen  in  inhaled  water:  squares,  pH  7.1;  solid  circles,  pH  7.9; 
and  crosses,  pH  8.7. 

DISCUSSION 

The  relationship  between  activity  and  metabolism  has  been  only  slightly  in- 
vestigated in  crustaceans.  In  one  study  of  the  mysicl.  My  sis  relicta,  it  has  been 
concluded  that  activity  during  vertical  migration  has  no  effect  on  metabolism 
(Foulds  and  Roff,  1975).  Another  study  concludes  that  oxygen  consumption  of 
euphausiids  remains  the  same  regardless  of  whether  the  animal  is  swimming  or  not 
(Lasker,  1966).  On  the  other  hand,  Halcrow  and  Boyd  (1967)  found  a  linear 
relationship  between  oxygen  consumption  and  swimming  activity  in  the  amphipod 
Gannnarus  occanicns,  and  Ivlev  (1963)  found  a  semi-logarithmic  relationship  be- 
tween oxygen  consumption  and  swimming  velocity  in  the  shrimp  Leander  adspcrsns. 
The  data  collected  in  this  study  show  a  quite  significant  relationship  between 
activity  and  metabolic  rate  in  Gnathophausia  hu/cns.  The  exponential  relationship 
between  rate  of  pleopod  beat  and  oxygen  consumption  in  this  species  is  comparable 
to  that  found  for  the  relationship  between  rate  of  cirral  beat  and  oxygen  consump- 
tion in  the  barnacle  Balanns  balanoides  (Newell  and  Northcroft,  1965).  Clarifica- 
tion of  the  nature  of  the  relationship  between  swimming  velocity  and  oxygen  con- 
sumption in  G.  ingcns  awaits  the  determination  of  the  propulsive  efficiency  of  the 
pleopod  beat.  It  is  clear  from  this  study  however,  that  the  rate  of  aerobic  metab- 
olism in  this  species  is  quite  closely  related  to  its  activity  level.  Furher,  activity 
certainly  constitutes  a  major  fraction  of  the  overall  energy  usage  of  this  entirely 
pelagic  species. 


146  T.  J.  MICKKL  AND  J.  J.  CHILDRESS 


At  the  relatively  high  activities  exhibited  by  individuals  during  experiments, 
respiratory  rates  could  range  from  30  to  80  p\  O-/(g  wet  weight -hr).  Whether 
these  rates  of  activity  can  be  maintained  in  the  oxygen  minimum  layer  can  be  calcu- 
lated by  assuming  an  oxygen  concentration  of  0.25  ml/liter  and  a  ventilation  flow 
rate  of  "240  ml/(g  wet  weight  -hr)  (Childress,  W71  ).  With  O2  extraction  of  85c/< , 
individuals  of  (/'.  ini/cns  can  extract  enough  oxygen  to  sustain  rates  of  activity  of 
138  and  172  pleopod  beats  per  minute  at  pH  7.1  and  pH  7.9. 

The  most  striking  aspect  of  the  data  on  this  species  is  that  there  are  no  dramatic 
effects  of  pH  on  its  respiratory  processes.  That  is,  this  shrimp  seems  to  regulate 
its  metabolism  over  a  wide  range  of  pH  values.  For  example,  the  relationship 
between  activity  and  oxygen  consumption  is  unaffected  by  pH.  Further,  pH  ap- 
pears to  have  little  effect  on  activity  at  the  two  lower  and  environmentally  more 
realistic  values.  Level  of  pH  also  apparently  does  not  alter  the  ability  of  this 
species  to  regulate  its  oxygen  consumption  at  the  low  environmental  oxygen  con- 
centrations W7here  it  normally  lives.  This  is  shown  by  the  fact  that  the  relationship 
between  critical  partial  pressure  and  regulated  oxygen  consumption  rate  found  by 
Childress  (1971)  is  unaffected  by  pH  over  the  range  tested. 

The  one  case  where  pH  had  a  strong  effect  involved  the  extraction  of  oxygen 
from  the  respiratory  stream.  The  data  show  quite  clearly  that  the  percentage  of  C>2 
extraction  is  10%  to  30%  higher  at  pH  7.1  as  compared  to  7.9  and  8.7.  Since  the 
utilization  at  pH  7.1  is  elevated  at  both  high  and  low  oxygen  partial  pressures,  this 
is  apparently  not  the  result  of  a  stress  on  the  oxygen  uptake  systems  of  the  animal 
forcing  it  to  increase  its  extraction  to  maintain  a  constant  oxygen  consumption. 
Paradoxically  this  higher  percentage  of  extraction  at  low  pH  apparently  does  not 
improve  the  ability  of  G.  inc/cns  to  regulate  its  oxygen  uptake  (Fig.  4).  This  im- 
plies that  there  is  a  loss  in  effectiveness  of  uptake  at  some  other  site  as  a  result  of  the 
lower  pH.  The  question  of  how  the  percentage  of  extraction  of  this  species  can 
be  higher  at  limiting  oxygen  partial  pressures  at  pH  7.1  as  compared  to  pH  7.9  and 
pH  8.7,  however,  is  still  left  unanswered.  Further  studies  on  the  tolerance  of 
vertically-migrating  animals  to  low  pH  mav  be  interesting  because  pH  may  be  as 
important  a  factor  as  oxygen  in  limiting  the  distribution  of  some  species  in  relation 
to  low  oxygen  regions. 


This  research  was  supported  by  NSF  grants  GA33232  and  OCE76-10407,  and 
administered  by  the  Marine  Science  Institute.  The  animals  used  in  these  studies 
were  captured  from  the  NSF  funded  vessels  R.  v.  VKLERO  iv  and  R.  v.  AGASSIZ.  We 
thank  L.  B.  Ouetin  and  J.  Torres  for  critically  reviewing  this  manuscript. 

SUMMARY 

1.  Pleopod  beat  of  G.  hu/cns  was  unaffected  by  pH  at  pH  7.1  and  pH  7.9  but 
was  lower  at  pH  8.7  due  to  increased  cleaning  activity. 

2.  The  relationship  between  oxygen  consumption  rate,  and  pleopod  beat  was 
found  to  be  semi-logarithmic. 

3.  The   relationship   between   oxygen   consumption   rate   and   activity   was   un- 
affected by  pH  in  the  range  of  pH  7.1  and  pH  8.7. 


EFFECT  OF  pH   IX   GNATHOPHAUSIA  147 

4.  The  per  cent  O2  extraction  of  oxygen  by  G.  ingcns  was  not  statistically  dif- 
ferent at  pH  7.9  and  pH  8.7,  but  was  greater  at  pH  7.1. 

5.  The  ability  of  G.  ingcns  to  regulate  its  oxygen  consumption  was  unaffected  by 
pH  in  the  range  studied. 

6.  Since  the  increase  in  per  cent  ()L.  extraction  at  pH  7.1  does  not  improve  the 
ability  of  G.  ingcns  to  regulate  its  oxygen  uptake,  it  appears  that  there  is  a  loss  in 
effectiveness  elsewhere  in  its  respiratory  system  at  this  pH. 

LITERATURE  CITED 

BANSE,  K.,  1964.     On  the  vertical  distribution  of  zooplankton  in  the  sea.     Pages  53-125  in  M. 

Sears,  Ed.,  Progress  in  oceanography,  Volume  II.    Pergamon  Press,  Oxford. 
BELMAN,   B.  W.,  AND  J.  J.  CHILDRESS,   1976.     Circulatory  adaptions  to  the  oxygen  minimum 

layer  in  the  bathypelagic  mysid  Gnathophausia  ingens.     Biol.  Bull.,  150:   15-37. 
CHILDRESS,  J.  J.,   1968.     Oyxgen  minimum  layer:    vertical   distribution  and  respiration  of  the 

mysid  Gnathophausia  ingcns.    Science,  160  :   1242-1243. 
CHILDRESS,   J.    J.,    1969.      The    respiratory    physiology    of   the    oxygen    minimum    layer    mysid 

Gnathophausia    ingcns.    Ph.D.    thesis.   Stanford    University,    Stanford,    142    pp.    (Diss. 

Abstr.,  30:  1271-B  ;  order  no.  69-13,934.) 
CHILDRESS,  J.  J.,  1971.     Respiratory  adaptions  to  the  oxygen  minimum  layer  in  the  bathypelagic 

mysid  Gnathopliaitsia  ingens.    Biol.  Bull.,  141  :   109-121. 
CHILDRESS,  J.  J.,  1975.     The  respiratory  rates  of  midwater  crustaceans  as  a  function  of  depth  of 

occurrence  and  relation  to  the  oxygen  minimum  layer  off  Southern  California.    Comp. 

Biochcm.  Physiol.,  SO  :  787-799. 
CLARK,   L.   C.,   1956.     Monitor  and  control   of  blood  and   tissue  oxygen  tensions.     Trans.  Am. 

Soc.  Artif.  Intern.  Organs,  2:  41-48. 
FOULDS,  J.  B.,  AND  J.  C.  ROFF,  1975.     Oxygen  consumption  during  simulated  vertical  migration 

in  Mysis  rclicta  (Crustacea,  Mysidacea).     Can  J.  Zoo!.,  54:  377-385. 
HALCROW,  K.,  AND  C.  M.  BOYD,  1967.     The  oxygen  consumption  and  swimming  activity  of  the 

amphipod   Gaiiniianis  oeeaniciis  at   different   temperatures.      Comf>.   Biochcm.   Physiol., 

23 :  233-242. 
IVLEV,  V.  S.,  1963.     Consumption  of  energy  during  movement  of  shrimps.     Zool.  Zh.,  42:  1465- 

1471  (in  Russian) . 

LASKER,  R.,  1966.     Feeding,  growth,  respiration,  and  carbon  utilization  of  a  euphausiid  crusta- 
cean.   /.  Fish.  Res.  Board  Can.,  23:  1291-1317. 
NEWELL,  R.  C.,  AND  H.  R.  NORTHCROFT,   1965.     The  relationship  between  cirral  activity  and 

oxygen  uptake  in  Balanus  balanoides.     J.  Mar.  Biol.  Assoc.  U.K.,  45:  387-403. 
PARK,  K.,  1968.     Alkalinity  and  pH  off  the  coast  of  Oregon.     Deep  Sea  Res.,  15:   171-183. 
QUETIN,  L.  B.,  AND  J.  J.   CHILDRESS,   1976.     Respiratory  adaptations  of  Plcuroncodes  planipcs 

Stimpson  to  its  environment  off  Baja  California.     Mar.  Biol.,  38:  327-334. 
QUETIN,  L.  B.,  T.  J.  MICKEL,  AND  J.  J.  CHILDRESS,  1978.     A  method  for  simultaneously  measur- 
ing the  oxygen  consumption  and  activity  of  pelagic  crustaceans.     Coinp.  Biochem.  and 

Physiol.,  in  press. 
SCHMIDT,  J.,  1925.     On  the  contents  of  oxygen  in  the  ocean  on  both  sides  of  Panama.    Science 

61 :  592-593. 
SEWELL,  R.  B.  S.,  AND  L.  FACE,  1948.    Minimum  oxygen  layer  in  the  ocean.    Nature,  162:  949- 

951. 
SMITH,  K.  L.,  JR.,  AND  J.  TEAL,   1973.     Temperature  and  pressure  effects  on  respiration  of 

thecosomatus  pteropods.     Deep  Sea  Res.,  20 :  853-858. 
TEAL,  J.,   1971.     Pressure  effects  on  the  respiration  of  vertically  migrating  decapod  Crustacea. 

Am.  Zool.,  11:  571-576. 
TEAL,  J.  M.,  AND  F.  G.  CAREY,  1967.     Respiration  of  a  euphausiid  from  the  oxygen  minimum 

layer.     Liinnol.  Occanogr.,  12  :  548-550. 


Reference:   B'wl.  Bull.,  154:  148-156.    (February,  1978) 


SEPARATION  AND   PARTIAL  PURIFICATION  OF   CENTRAL  NER- 
VOUS SYSTEM  PEPTIDES  FROM  LIMULUS  POL  YPHEMUS  WITH 
HYPERGLYCEMIC  AND  CHROMATOPHOROTROPIC  ACTIVITY 

IN  CRUSTACEANS  '• : 

PAUL  D.  PEZALLA,  ROBERT  M.  DORES,  AND  WILLIAM  S.  HERMAN  3 

Department  of  Genetics  and  Cell  Hiohn/y,  University  of  Minnesota,  St.  Paul,  Minnesota  55108 

Thirty-six  years  ago  a  substance  in  the  central  nervous  system  (CNS)  of  the 
chelicerate  arthropod  Limit! us  poIypJicmus  was  shown  to  possess  chromatophoro- 
tropic  activity  when  tested  on  the  mandihulate  arthropod  Uca  puyna.v  (Brown  and 
Cunningham,  1941).  More  recent  studies  have  demonstrated  that  CNS  extracts 
from  Li nnil us  are  also  chromatophorotropic  in  a  variety  of  other  decapods,  including 
both  brachyuran  and  natantian  species  (Fingerman,  Bartell,  and  Krasnow,  1971  ; 
Herman  and  Dallmann,  1975).  Other  experiments  have  shown  that  arthropod 
molting  hormones  (ecydsones)  are  present  and  active  in  Liniuliis  (Krishnakumaran 
and  Schneiderman,  1970;  Jegla,  Costlow  and  Alspaugh,  1972;  Winget  and  Herman, 
1976).  The  existence  of  both  CNS  material  with  crustacean  neurosecretory  hor- 
mone activity  and  ecdysones  in  this  species  suggests  that  Liuntlns  might  also  pro- 
duce other  substances  with  arthropod  hormone  activity.  If  so,  neuroendocrinologi- 
cal  studies  of  Limn! its  could  be  of  major  importance  in  attempts  to  understand  the 
basic  properties  and  evolution  of  arthropod  neuroendocrine  regulatory  mechanisms. 
Against  this  background  a  series  of  studies  were  conducted  testing  the  effects  of 
CNS  extracts  from  Liiiiitlns  in  known  arthropod  neurosecretory  hormone  bioassays. 
During  this  work  the  existence  of  a  CNS  substance  causing  hyperglycemia  in  the 
freshwater  crayfish,  Orconcctcs  imnninis,  was  discovered.  Initial  studies  on  this 
substance,  and  evidence  demonstrating  that  it  is  not  the  above-mentioned  chromato- 
phorotropin,  are  presented  below. 

MATERIALS  AND  METHODS 

Adult  specimens  of  Limit! its  [>olyplicmns,  obtained  from  the  Marine  Biological 
Laboratory,  \Yoods  Hole,  Massachusetts,  were  maintained  without  feeding  in  In- 
stant Ocean  artificial  sea  water  at  12°  C.  Specimens  of  Orconcctcs  immitnis,  ob- 
tained from  Trans-Mississippi  Biological  Supply,  St.  Paul,  Minnesota,  were  main- 
tained in  dechlorinated  tap  water  aquaria  at  12°  C  and  fed  Gainesburger  dog  food 
three  or  four  times  a  week.  Specimens  of  Uca  pityilator,  supplied  by  Gulf  Speci- 
men Co.,  Panacea,  Florida,  were  held  in  Instant  Ocean  sea  water  aquaria  at  18°  C 
and  fed  Gainesburger  dog  food  weekly.  Eyestalks  were  removed  from  specimens  of 
Orconcctcs  and  Uca  24  hr  prior  to  experiments. 

Central  nervous  systems  from  Limn! its  were   removed  by  ventral   dissection, 

1  Supported  by  the  University  of  Minnesota  Graduate  School  and  USPHS  grant  HD-07336. 

2  Some  of  this  research  was  part  of  a  Ph.D.  thesis  submitted  by  P.  D.  Pezalla  to  the  Uni- 
versity of  Minnesota. 

3  To  whom  reprint  requests  should  be  sent. 

148 


CNS  PEPTIDES   FROM   l.IMULUS  149 

cleaned  of  adhering  non-CNS  material,  weighed  to  the  nearest  nig  and  either  im- 
mediately homogenized,  or  frozen  on  dry  ice  for  lyophilization.  Extracts  were 
prepared  with  ethanol,  acetone,  0.1  and  1.0  x  acetic  acid,  and  0.1  M  ammonia.  The 
typical  extraction  protocol  was  as  follows.  The  CXS  was  placed  in  a  volume  of 
solvent  and  thoroughly  homogenized  in  a  Potter  Elvejem  homogenizer  at  4°  C. 
The  homogenate  was  then  centrifnged  for  15  min  at  12,100  X  g  in  a  Sorval 
Superspeed  RC2-B  refrigerated  centrifuge  run  at  4°  C.  The  supernatant  was 
saved,  while  the  pellet  was  re-extracted  three  to  four  times  to  a  total  of  20  volumes. 
The  pooled  supernatants  were  then  hoiled  for  three  min  and  recentrifuged  as 
mentioned  above.  The  crude  extract  was  either  used  immediately  or  lyophilized 
for  storage  at  -—20°  C.  In  most  cases,  the  lyophilized  residues  were  redissolved 
in  distilled  water  at  concentrations  appropriate  for  injections  or  column  chromatog- 
raphy.  In  some  experiments  crude  extract  was  made  10~3  M  with  thiodiglycol 
(Sigma).  Variations  from  the  above  procedure  are  cited  in  the  text. 

Assays  of  the  CNS  chromatophorotropin  from  Liinulns,  hereafter  referred  to  as 
LUC,  were  conducted  as  previously  described  (Herman,  1975).  In  brief,  the 
melanophore  response  of  5-10  eyestalkless  female  Uca  to  10  /J  aliquots  of  extract 
or  solvent  was  observed,  the  control  values  subtracted  from  the  experimental  values, 
and  the  mean  response  per  Uca  calculated  in  chromatophore  units. 

Assays  of  CNS  hyperglycemic  activity  from  Liinulns  were  performed  on  eye- 
stalkless, mixed  sex  Orconec.tes  randomly  assigned  to  individual  containers  holding 
enough  dechlorinated  tap  water  to  just  cover  the  carapace.  Injections  of  50-100 
ju.1  of  solvent  or  extract  were  made  with  disposable  1  ml  tuberculin  syringes  fitted 
with  25  gauge  needles.  Hemolymph  samples,  withdrawn  from  the  ventral  abdominal 
or  cephalothoracic  sinus,  were  assayed  for  glucose  by  the  Glucostat  (Worthington 
Biochemicals)  method  (Meites.  1965)  or  for  total  carbohydrate  by  the  Anthrone 
(Sigma)  method  (Chaykin.  1966).  A  Beckman  DB-G  spectrophotometer,  set  at 
540  nm  for  Glucostat  and  620  nm  for  Anthrone,  was  used  for  all  colorimetric 
determinations.  Glucose  was  used  as  the  standard  for  both  assay  procedures. 

Chromatography  of  LUC  and  experiments  concerning  both  LUC  and  the  hyper- 
glycemic factor  were  conducted  at  4°  C  on  a  Sephadex  G-25  Fine  (Pharmacia) 
column,  1.5  X  90  cm,  equilibrated  with  1.0  N  acetic  acid  made  10"3  M  with  thiodi- 
glycol. The  flow  rate  was  25  ml/hr  and  2.5  ml  fractions  were  collected.  Absorb- 
ancy  of  all  fractions  was  read  at  280  nm.  The  fractions  were  lyophilized  and  redis- 
solved in  distilled  water.  The  column  was  calibrated  with  lysozyme  (14,000), 
ACTH  (4,570),  glucagon  (3,600)  and  bacitracin  (1,400),  all  from  Sigma. 

The  hyperglycemic  factor  was  further  chromatographed  on  a  Sephadex  G-50 
Fine  (Pharmacia)  column,  2  X  40  cm,  equilibrated  with  0.1  N  acetic  acid  made  10"3 
M  with  thiodiglycol.  This  column  was  run  at  4°  C  with  a  flow  rate  of  4.6  ml/hr, 
and  fractions  of  4.6  ml  were  collected.  The  fractions  from  this  column  were  treated 
as  previously  mentioned  with  the  exception  that  in  some  experiments  the  lyophilized 
fractions  were  redissolved  in  0.1  N  acetic  acid.  (Several  experiments  demon- 
strated that  50  pi  of  this  solvent  were  not  hyperglycemic  in  crayfish  and  thus  did  not 
interfere  with  the  hyperglycemic  assay.)  This  column  was  calibrated  with  bovine 
serum  albumin  (68,000),  chymotrvpsinogen  (25,000),  lysozyme  (14,000)  and 
glucagon  (3600),  all  from  Sigma. 

LUC  and  the  hyperglycemic  factor  were  tested  for  susceptibility  to  some  or  all 


150  PEZALLA,  DORES  AND  HERMAN 

TAHLE     I 

Effect  of  CNS  extracts  from  Liimilus  on  Orconertes  hemolymph  glucose. 

Hemolympli  glucose 
Material  injected  (mg  %) 

CNS  equivalents 

0.04  (8)  8.7  ±  2.3 

0.08  (4)  12.1  ±  4.4 

0.20  (17)  19.2  ±  2.2 

Solvent  control  (14)  3.6  ±  0.4 

N  in  parentheses;  experimental  duration  =  1  hr. 

of  the  following  enzymes  :  pepsin,  protease,  trypsin,  chymotrypsin,  thermolysin,  and 
lysozyme  (all  from  Sigma). 

For  each  enzyme,  extracts  were  incubated  in  the  appropriate  medium  (enzyme 
concentration  —  20  mg/ml)  for  16  hr  at  37°  C.  The  reactions  were  terminated  by 
boiling  the  reaction  mixture  for  5  min,  after  which  the  reaction  mixtures  were 
centrifuged  to  remove  denatured  enzyme.  Extract  without  enzyme  and  enzyme 
without  extract  controls  were  subjected  to  the  same  conditions.  The  following 
buffers  were  used  (Shepard,  1975)  :  pepsin,  0.1  N  acetic  acid  (pH  2.8)  ;  protease, 
0.02  M  HEPES-KOH  (pH  7.5)  containing  0.1  M  calcium  chloride;  trypsin,  0.05 
M  Tris  (pH  8.2)  containing  0.01  M  calcium  chloride;  chymotrypsin,  0.08  M  Tris 
(pH  7.8)  containing  0.1  M  calcium  chloride;  thermolysin,  0.5  M  Tris  (pH  8.5) 
containing  0.005  M  calcium  chloride;  and  lysozyme,  0.1  M  phosphate,  pH  7.0. 

The  data  are  reported  as  mean  ±  s.e.m.  Some  of  these  data  were  analyzed  by 
Student's  /-test ;  the  term  significance  in  this  report  refers  to  statistical  significance 
in  this  test  at  the  5(/o  level  or  better. 

RESULTS 
Effects  of  CNS  extracts  from  Limulus  on  Orconectes  hemolymph  carbohydrates 

Initial  studies  tested  CXS  extracts  from  Li  in  nl  its  for  hyperglycemic  activity  in 
Orconectes.  The  results  of  a  typical  experiment,  using  acetone  extracts  of  fresh 
Limulus  CNS,  are  summarized  in  Table  I.  Significant  elevations  of  hemolymph 
glucose  were  obtained  with  as  little  as  0.04  CXS  equivalent/crayfish,  and  larger 
doses  produced  substantially  higher  reponses.  Analysis  of  total  hemolymph  car- 
bohydrate before  and  after  injection  yielded  similar  results;  in  one  such  experiment 
Limulus  CNS  extracts  (0.20  CNS  equivalent/animal)  produced  a  130.0  ±  17.3% 
increase  in  10  crayfish,  while  injections  of  solvent  or  muscle  extract  into  20  animals 
elevated  total  carbohydrate  by  only  32.0  ±  7.57' •  Comparable  experiments  have 
been  performed  several  times  over  a  period  of  2  yr  using  horseshoe  crabs  and  cray- 
fish obtained  in  both  summer  and  winter.  These  studies  invariably  demonstrated 
that  Limn/its  CNS  extracts  contained  material  capable  of  rising  Orconectes  hemo- 
lymph total  carbohydrate  and  glucose  levels. 

The  above  results  are  duplicated  with  acetic  acid  extracts  of  Liinnlits  CNS; 
ethanol  and  ammonia  extracts  also  cause  significant  responses,  but  hemolymph 
glucose  increases  are  quantitatively  less  impressive.  Hyperglycemic  activity  can 


CNS  PEPT1DES  FROM  LIMULUS 


151 


40         50 


60 


70         80        90        100       110 
ELUTION    VOLUME    (ml  ) 


120 


FIGURE  1.  Chromatography  of  crude  CNS  acetic  acid  extracts  from  Limulus  on  Sephadex 
G-25  (fine).  The  extracts  consisted  of  3500  nig  of  CXS  ( \vet  weight)  extracted  as  described 
and  concentrated  to  3  ml.  The  column  was  equilibrated  with  1.0  x  acetic  acid;  flow  rate,  25 
ml/hr;  fraction  volume,  4  nil;  total  volume,  130  ml;  void  volume,  54  ml.  The  absorbancy  of 
each  fraction  was  read  at  280  nm  (solid  line).  The  fractions  were  pooled  and  assayed  for 
hyperglycemic  and  melanophore  dispersing  activity  (Table  II). 

usually  be  extracted  from  fresh  or  lyophilized  CXS,  but  acetone  extracts  of  lyo- 
philized material  have  no  effect.  For  convenience  we  have  designated  the  active 
material  in  CXS  extract  LHGF,  for  Liinnlns  hyperglycemic  factor. 

Separation  of  LHGF  and  Li'C  on  Scphatic.v  G-25 

Preliminary  gel  filtration  experiments  (Fingerman  ct  <//.,  1971  )  indicated  that 
LUC  would  elute  in  the  last  half  of  the  elution  volume  on  a  Sephadex  G-25  column. 
\Ye  therefore  decided  to  attempt  to  separate  LHGF  and  LUC  by  means  of  such  a 
column.  Concentrated  Liiunlits  CXS  acetic  acid  extract  was  applied  to  the  column, 
and  4  ml  fractions  were  collected,  pooled  (as  indicated  in  Fig.  1  ).  lyophilized,  and 
redissolved  in  distilled  water  for  bioassay.  The  results  were  obvious  (see  Table 
II )  ;  LHGF,  but  not  LUC,  was  present  in  Fraction  I,  which  corresponded  to  the 
void  volume.  Xeither  activity  was  present  in  Fraction  II,  while  Fraction  III  ex- 
hibited only  LUC  activity.  From  these  experiments  it  was  concluded  that  LHGF 
is  excluded  from  Sephadex  G-25  columns. 

Preliminary  characterization  of  LHGF 

The  preceding  results  clearly  indicated  that  LHGF  and  LUC  were  separate 
substances.  It  was  therefore  decided  to  further  characterize  both  LHGF  and  LUC 
to  demonstrate  their  nonidentity. 

Experiments  were  undertaken  to  determine  the  stability  of  LHGF.  It  was  ob- 
served that  unboiled  acetic  acid  extracts  of  Liinii/ns  CXTS  are  unstable  at  room 
temperature,  with  the  majority  of  the  hyperglycemic  activity  lost  within  3  hr.  This 
loss  in  activity  could  be  prevented  by  brief  boiling  of  the  crude  extract,  or  In- 
storing  the  crude  extract  of  0°  C.  In  addition,  treatment  of  the  crude  extract  with 
hydrogen  peroxide  (final  concentration  =:  1^)  reduced  hyperglycemic  activity  by 


152 


PEZALLA,  DORES  AND  HERMAN 


TABLE  II 
Effects  of  Srphadex  G-25  fractions  of  CNS  extracts  from  Limulus  in  Orconectes  and  Uca. 


Fraction  tested 


Uca  response* 


Orconectes  response** 


I 

II 

III 

Solvent  controls 


3.0  ±  1.8  (10) 

1.6  ±  1.1  (10) 

23.0  db  3.0  (10) 

0.0  (10) 


86.0  ±  13.0  (10) 
17.0  ±  6.0  (10) 
22.5  ±  3.0  (10) 
16.5  ±  4.5  (10) 


*  Mean  net  chromatophore  response;  10  n\  of  pooled  fraction  injected. 

**  Percentage  of  increase  in  total  carbohydrate  60  min  after  injection;  100  n\  of  pooled  fraction 
injected. 

about  one-third  in  1  hr.  On  the  basis  of  the  above  findings  we  now  routinely  boil 
the  crude  extracts  briefly,  centrifuge,  and  add  thiodiglycol  to  a  final  concentration 
of  10"3  M.  In  addition,  all  extractions  and  chromatographic  separations  are  per- 
formed at  4°  C 

The  next  concern  was  to  estimate  the  molecular  weight  of  the  LHGF  via  gel 
filtration.  Initial  separation  of  LHGF  and  LUC  on  Sephadex  G-25  clearly  demon- 
strated that  a  larger  grade  Sephadex  was  required ;  G-50  was  selected.  The  results 


o 

CO 
CO 


co 

^. 

UJ 


o 

r- 
Q_ 

o 


2.0 


1.0 

0  8 
0.6 

04 
0.2 


Chymotrypsmogen 
(  2  5,000) 

.  Lysozyme 
(14,000) 


0  I 


Kav 


/  \ 


LJ 
CO 
O 

O 

ID 
_l 
CD 


CP 

£ 


46 


69  92  115 

ELUTION  VOLUME   (ml) 


138 


FIGURE  2.  Chromatography  of  crude  CNS  acetic  acid  extracts  from  Limulus  on  Sephadex 
G-50  (fine).  The  column  was  equilibrated  with  0.1  N  acetic  acid;  flow  rate,  4.6  ml/hr;  frac- 
tion size,  4.6  ml;  total  volume,  140  ml;  void  volume,  51  ml.  The  absorbancy  of  each  fraction 
was  read  at  280  (dashed  line).  The  hyperglycemic  activity  of  each  fraction  was  tested  on 
three  animals  per  fraction  (solid  line).  Insert  shows  G-50  calibration  curve. 


CNS  PEPTIDES  FROM  LIMULUS 


153 


of  a  typical  G-50  experiment  are  shown  in  Figure  2.     On  this  column  LHGF 
eluted  as  a  single  symmetric  peak  with  an  estimated  molecular  weight  of  6400. 

Active  fractions  from  the  G-50  runs  were  next  treated  with  various  proteolytic 
enzymes.  Incubation  of  crude  extracts  with  pepsin  or  protease  resulted  in  a  de- 
crease in  LHGF  activity  of  92.4%  and  79.7%,  respectively,  while  incubation  with 
trypsin  had  no  effect  on  activity. 

Preliminary  characterisation  of  LUC 

Several  experiments  were  conducted  on  LUC  to  determine  its  stability,  suscepti- 
bility to  various  proteolytic  enzymes,  and  molecular  weight.  As  reported  elsewhere 
(Brown  and  Cunningham,  1941  ),  brief  boiling  of  crude  extracts  had  no  effect  on 
activity.  However.  LUC  did  appear  to  be  susceptible  to  oxidation.  The  combined 
results  of  experiments  conducted  at  both  20°  C  and  37°  C  showed  that  untreated 
extracts  held  for  20  hr  lost  32.9%  of  the  original  activity.  In  addition,  treatment 
with  thiodiglycol  prevented  this  loss,  while  treatment  with  hydrogen  peroxide  lead 
to  a  63.9%  loss  in  activity.  In  view  of  these  results,  all  subsequent  extracts  were 


o 
oo 

00 


CO 

-z. 

LU 
O 


2  0- 


I  0- 


0.8- 


o 

JZ     06 

Q_ 

0    04 


0.2 


ACTH    (4600) 

Glucogon 
(3600) 


LUC 
(1850) 


l\ 


i    I 


O.I 


Kav 


1.0 


20 


LU 
CO 

15  1 

CO 
UJ 

a: 


10 


20 


40 


60 


80 


100 


120 


ELUTION    VOLUME    (ml) 


LL) 


Q_ 
O 


o 
cr 

x 
o 


FIGURE  3.  Chromatography  of  crude  CNS  acetic  acid  extract  from  Limulns  on  Sephadex 
G-25  (fine).  The  column  was  equilibrated  with  1.0  x  acetic  acid;  flow  rate,  25  ml/hr;  frac- 
tion volume,  2.5  ml;  total  volume,  134  ml;  void  volume,  46  ml.  The  crude  extract  consisted 
of  350  mg  CNS  (wet  weight)  extracted  as  described  and  concentrated  to  2.5  ml.  The  absor- 
bancy  of  each  fraction  was  measured  at  280  nm  (dashed  line).  The  chromatophorotropic  activ- 
ity of  each  fraction  was  tested  on  18  animals/fraction  and  the  response  depicted  (solid  line). 
Insert  shows  G-25  calibration  curve. 


154  I'EXALLA,  DORKS   AND   1IKKMAN 

made  10  :;  M  with  thiodiglycol.  Treatment  of  crude  Limiting  CXS  extracts  with 
various  enzymes,  namely  protease,  pepsin,  chymotrypsin,  trypsin,  and  thermolysin 
resulted  in  a  mean  decrease  in  I  AX'  activity  of  92.8  ±  l.4r/( ,  while  the  glycosidase 
lysozyme  was  without  effect. 

Molecular  weight  determination  was  done  on  a  calibrated  Sephadex  G-25 
column.  As  indicated  in  Figure  3,  a  major  peak  of  activity  eluted  with  an  esti- 
mated molecular  weight  of  1,850  trailed  by  a  secondary  peak  of  activity.  The 
presence  of  this  latter  peak,  also  noted  by  Fingerman  ct  al.  (1971),  suggests  the 
possible  existence  of  more  than  one  substance  with  LUC  activity  in  the  CXS  of 
Limulus. 

DISCUSSION 

These  experiments  have  demonstrated  the  existence  of  two  dissimilar  sub- 
stances with  hormonal  activity  in  crustaceans  in  Limnlns  polyphemus  CNS  extracts. 
One  of  these  substances,  the  previously  unreported  LHGF,  apparently  has  a 
molecular  weight  of  about  6.400  daltons.  In  addition,  it  appears  to  be  heat  stable, 
inactivated  by  hydrogen  peroxide,  sensitive  to  some  proteolytic  enzymes,  and  un- 
affected by  incubation  with  trypsin.  These  data  collectively  indicate  that  LHGF 
is  a  polypepticle.  LHGF  is  clearly  hyperglycemic  in  Orconcctcs,  but  lacking  in 
melanophore  pigment  dispersing  activity  in  Uca.  The  second  substance (s)  is  the 
previously  known  chromatophorotropin,  LLTC.  These  studies  have  added  to  exist- 
ing knowledge  of  this  substance  by  demonstrating  an  apparent  molecular  weight  of 
1850  daltons,  an  estimate  in  agreement  with  earlier  studies  (Fingerman  ct  al., 
1971).  In  addition,  it  has  been  shown  for  the  first  time  that  LUC  is  susceptible  to 
a  variety  of  proteolytic  enzymes,  including  trypsin,  and  to  the  oxidizing  agent 
hydrogen  peroxide.  Previous  reports  have  demonstrated  that  LUC  is  heat  stable 
(Brown  and  Cunningham,  1941  ;  Herman,  1975).  The  total  available  data  indicate 
that  LLTC  is  a  peptide ;  it  is  chromatophorotropic  in  several  crustaceans  (see  Her- 
man and  Dallman,  1975),  but  lacks  hyperglycemic  activity  in  Orconectcs.  On  the 
basis  of  the  above  results,  it  can  be  concluded  that  Liniulits  CNS  extracts  contain  at 
least  two  distinct  peptides  with  different  hormonal  activity  in  crustaceans. 

These  data  provide  a  basis  for  comparison  of  the  properties  of  LHGF  and  LUC 
with  those  of  known  crustacean  neurosecretory  hormones.  The  available  biological 
and  chemical  evidence  indicates  the  existence  of  more  than  one  decapod  hyper- 
glycemic hormone  (Kleinholz  and  Keller,  1973;  Kleinholz,  1976).  Studies  to 
determine  the  interspecific  effect  of  various  decapod  hyperglycemic  hormones 
(Keller,  1969)  indicate  little  cross  reactivity  among  the  major  suborders  of  deca- 
pods. However,  partial  chemical  characterization  of  the  hyperglycemic  hormones 
from  Cancer  magistcr,  Pandalits  jordani  and  Orconcctcs  litnosns  suggests  chemical 
similarity ;  i.e.,  all  appear  to  have  molecular  weights  of  about  7,000  daltons,  all  are 
heat  labile  and  susceptible  to  some  proteolytic  enzymes,  and  at  least  some  appear 
to  be  resistant  to  trypsin  and  inactivated  by  hydrogen  peroxide  (Kleinholz,  Kim- 
ball  and  McGarvey,  1967;  Kleinholz  and  Keller,  1973;  Kleinholz,  1976).  LHGF 
also  appears  to  fit  into  this  basic  scheme,  with  the  notable  exception  that  it  is  ap- 
parently heat  stable.  Decapod  melanophore  pigment  dispersing  hormones  (MDH) 
currently  seem  to  be  less  heterogeneous  than  the  hyperglycemic  hormones ;  they  ap- 
pear to  be  biologically  similar,  heat  stable,  susceptible  to  proteolytic  enzymes  and 


CNS  PEPTIDES  FROM  LIMULUS  155 

oxidation,  and  to  have  molecular  weights  of  about  2,000  daltons.  The  studies  of 
Kleinholz  (1976)  suggest  that  the  MDHs  may  possess  a  structure  comparable  to 
that  of  the  distal  retinal  pigment  hormone  characterized  by  Fernlund  (1976).  The 
existing  data  therefore  suggest  that  LUC  and  LHGF  both  resemble  known  crusta- 
cean hormones  of  comparable  biological  activity.  Unfortunately,  since  only  the 
distal  retinal  pigment  hormone  of  Pandalits  borcalis  has  been  totally  characterized 
(Fernlund,  1976),  the  question  of  the  identity,  or  nonidentity,  of  these  various 
molecules  will  not  be  resolved  until  more  complete  structural  data  are  available. 

While  the  activity  of  LHGF  and  LUC  in  decapods  is  evident,  the  role  of  these 
peptides  in  Limn  Ins  remains  enigmatic.  LUC  cannot  act  on  integumentary  chroma- 
tophores  in  the  horseshoe  crab,  since  this  species  lacks  such  chromatophores. 
Similarly,  we  have  been  unable  in  several  attempts  to  demonstrate  a  hyperglycemic 
effect  of  partially  purified  LHGF  in  Limnlus. 

It  is  becoming  more  evident  that  the  neuroendocrine  system  of  Limnlus  is  de- 
serving of  further  study.  This  species  produces  and  uses  ecdysones  (Jegla  et  al., 
1972;  Winget  and  Herman,  1976),  and  it  possesses  at  least  two  neurosecretory 
hormone-like  peptides  active  in  mandibulates.  It  is  certainly  reasonable  to  expect 
that  further  studies  on  this  species  will  be  of  major  importance  in  our  attempts  to 
understand  the  neuroendocrinology  of  chelicerate  arthropods  and  the  evolution  of 
arthropod  neuroendocrine  systems. 


We  wish  to  express  our  appreciation  to  William  Sparkes  and  Louisa  Moore  for 
their  contributions  to  this  research. 

SUMMARY 

1.  Crude  extracts  of  Limulus  CNS  cause  hyperglycemia  in  Orconectcs  i  in  munis 
and  expand  chromatophores  in  Uca  pugilator. 

2.  The   hyperglycemic   action    is    due    to    a    previously    unknown    polypeptide 
(LHGF)    with  an  estimated  molecular  weight   of  6400  daltons.     LHGF   is   in- 
activated by  hydrogen  peroxide,  pepsin,  and  protease,  but  unaffected  by  trypsin 
and  brief  boiling. 

3.  The  chromatophorotropic  activity   is   due  to   the  previously   reported   sub- 
stance, LUC.    LUC  is  shown  to  be  a  peptide  with  an  approximate  molecular  weight 
of  1850  daltons;  it  is  inactivated  by  hydrogen  peroxide,  protease,  pepsin,  trypsin, 
chymotrypsin,  and  thermolysin. 

4.  LUC  and  LHGF  activity  can  be  readily  separated  by  gel  filtration  on  a 
Sephadex  G-25  column. 

5.  The  similarity  of  LLTC  and  LHGF  to  known  crustacean  hormones  is  dis- 
cussed. 

LITERATURE  CITED 

BROWN,  F.  A.,  JR.,  AND  O.  CUNNINGHAM,  1941.  Upon  the  presence  and  distribution  of  a 
chromatophorotropic  principle  in  the  central  nervous  system  of  Limulus.  Biol.  Bull.,  81 : 
80-95. 

CHAYKIN,  S.,  1966.    Biochemistry  laboratory  techniques.    Wiley  and  Sons,  New  York,  88  pp. 


156  PEZALLA,    DORKS   AND    IIKRM\\ 

FERNLUND,  P.,  1976.     Structure  of  light-adapting  hormone  troni  the  shrimp  /'nndtilus  bo 
Biocheui.  Biophys.  Ada,  439:  17-25. 

FINI.KKMAN,  M.,  C.  K.  BARTKU.,  AND  K.  A.  KKASNOXV,  1971.  Comparison  of  chromatophoro- 
tropins  from  tin-  horseshoe  crab  I.iinu/iis  f>ol\f>hciiins.  and  tin-  fiddler  crab,  i'cu  pm/i- 
lator.  Biol.  Hull..  140:  376-388. 

MIKMAN,  W.  S.,  1975.  Quantification  of  the  J.iiinilus  polyphemus  CX.S  chromatophorotropin. 
Gen.  Comp.  EnilocrinoL.  27  :  84-87. 

HERMAN,  W.  S.,  AND  S.  H.  DALLMANN,  1975.  Luiinlus  chromatophorotropin:  action  on  iso- 
lated Uca  legs  and  in  various  crustaceans.  E.vpcricntui.  31  :  918-919. 

JKC.LA,  T.  C.,  J.  D.  COSTLOW,  AND  J.  Ai.srAr<;n,  1972.  Effects  of  ecdysones  and  other  synthetic 
analogs  on  horseshoe  crab  larvae.  Cicn.  Coinfi.  EndocrinoL,  19:  159-167. 

KELLER,  R.,  1969.  Untersuchungen  r/.\\r  artspezifitat  eines  crustacean  hormones.  '/..  I'crc/l. 
/>/j.v.m./..63:  137-145. 

KLEINHOLZ,  L.  H.,  1976.  Crustacean  neurosecretory  hormones  and  physiological  >pecificity. 
Am.Zool.,  16:  151-167. 

KLEINHOLZ,  L.  H.,  AND  R.  KELLER,  1973.  Comparative  studies  in  crustacean  neurosecretory 
hyperglycemic  hormones  I.  The  initial  survey.  Gen.  Coiup.  EndocrinoL,  21  :  554-564. 

KLEINHOLZ,  L.  H.,  F.  KIMBALL,  AND  M.  McGARVEY,  1967.  Initial  characterization  and  separa- 
tion of  hyperglycemic  (diabetogenic)  hormone  from  the  crustacean  eyestalk.  Gen. 
Comp.  EndocrinoL,  8:  75-81. 

KRISHNAKUMARAN,  A.,  AND  H.  SCHNEIDERMAN,  1970.  Control  of  molting  in  mandibulate  and 
chelicerate  arthropods  by  ecdysones.  Hiol.  Bull.,  139:  520-538. 

MEITES,  S.  (Ed.),  1965.  Ultramicro  glucose  (enzymatic)  assay.  Pages  113-120  in  Standard 
methods  of  clinical  chemistry.  }'<>!.  5.  Academic  Press,  New  York. 

SHEPARD,  J.  G.,  1975.  A  polypeptide  sperm  activator  from  male  Saturniid  moths.  Insect 
Physio!.,  21 :  9-23. 

WINGET,  R.  R.  AND  W.  S.  HERMAN,  1976.  Occurrence  of  ecdysone  in  the  blood  of  the  cheli- 
cerate arthropod,  Limit/us  polyphcniiis.  E.rpcricntia,  32:  1345-1346. 


Reference:   Biol.  Bull..  154  :  157-175.    (February. 


DEVELOPMENT  OF  THE   EOLID  NUDIBRAXCH    CUTHONA    NANA 

(ALDER  AND  HANCOCK.  1842  i.  AND  ITS  RELATIONSHIP  WITH 

A  IIYDROII)  AXD  HERMIT  CKAI5 

BRIAX  R.  RIVEST  i 
Department  of  Zoology,  University  of  \'ei\.'  Hampsliire,  Ditiiuiiit,  AY<i'  Hampshire  1)3824 

Two  aspects  of  the  biology  of  the  eolid  nudibranch  Cnthoihi  nana  (  Alder  and 
Hancock,  1842)  are  examined  here.  The  development  of  C.  nana  \vas  studied  be- 
cause poecilogony  (different  developmental  patterns  within  a  species)  was  suspected. 
The  distribution  and  behavior  of  C.  nana  was  investigated  because  of  the  nudi- 
branch's  specialization  on  a  sedentary  prey  species  which  is  effectively  mobile  due 
to  its  commensal  relationship  with  hermit  crabs. 

In  1971  cultured  egg  masses  of  Cuthona  nana  developed  into  actively  swimming, 
planktotrophic  veligers ;  whereas  egg  masses  cultured  in  1(>73  produced  lecitho- 
trophic,  nonswimming  veligers  that  metamorphosed  within  a  day  or  two  of  hatching 
(Harris.  Wright,  and  Rivest.  1975).  Poecilogony  may  occur  in  some  opistho- 
branchs  (  Berrill.  1931  ;  Rasmussen.  11'44;  Franz,  1970.  1971),  but  this  phenomenon 
is  rare  among  marine  benthic  invertebrates  and  needs  to  be  investigated  further. 

In  the  study  reported  here,  egg  masses  cultured  initially  in  the  presence  of  the 
adult's  food,  Hydractinia  cch'niata  Fleming,  1828,  developed  only  in  the  nonpelagic 
lecithotrophic  mode.  In  an  attempt  to  induce  alternate  modes  of  development,  tem- 
perature, adult  nutrition  and  exposure  to  H.  cchinata  were  manipulated  on  different 
egg  masses  and  embrvogenesis  and  metamorphosis  were  followed.  Field  data  are 
compared  with  laboratory  observations. 

The  ecology  of  Cntlunut  nana  involves  a  species-specific  predator-prey  associa- 
tion with  the  colonial  hvdroid.  Hydractinia  ccJiinata,  commonlv  found  on  gastropod 
shells  occupied  by  pagurid  crabs  (  Fig.  1).  Hydractinia  cchinata  is  a  dioecious 
hvdroid  consisting  of  a  basal  mat  from  which  arise  gastrozooids,  gonozooids,  and 
defensive  dactylozooids  (  Hyman,  1940).  The  motile  nature  of  hermit  crabs  gives 
the  hydroid's  substrate  a  mobility  that  presents  potential  settlement  problems  for 
the  veligers  or  newly  metamorphosed  juveniles  of  C.  nana,  and  possible  prey-locat- 
ing difficulty  for  adult  nudibranchs.  Information  from  the  literature,  laboratory  and 
field  observations,  and  experiments  reveals  how  the  behavior  and  life  histories  of 
the  hvdroid,  nudibranch  and  hermit  crab  are  inter-related. 


.\l  ATKKIAI.S  AND  METHODS 

Specimens  of  Cutlnnia  nana,  their  egg  masses,  and  hermit  crabs  in  mollusc  shells 
bearing  Hydractinia  cchinata  were  collected  by  scuba  diving  in  Gosport  Harbor  at 
the  Isles  of  Shoals  (43°  59'  N;  70°  37'  W  )".  ca.  10  km  off  the  New  Hampshire 
coast.  Most  of  the  collecting  was  done  at  the  depth  of  3-12  m  in  Haley's  Cove,  an 

1  I're-ent  address:  Department  of  Zoology,  University  of  \Yashington,  Seattle,  Washington 
98195. 

157 


158  BRIAN  R.  RIVEST 

area  within  the  harbor  that  had  the  highest  concentrations  of  hermit  crabs  with  II. 
echinata  colonies.  Monthly  field  observations  and  collections  were  made  from 
January,  1974,  to  July,  1975,  excluding-  June  through  August,  1974.  The  hermit 
crabs,  hydroid  colonies,  and  nudibranchs  were  maintained  at  11-13°  C  in  a  re- 
circulating  seawater  system.  Within  two  days  of  their  collection,  the  colonies  of 
H.  echinata  were  examined  under  a  dissecting  microscope  and  the  numbers  and 
lengths  of  C.  nana  individuals  found  on  each  shell  were  recorded.  The  ciliated 
epithelium  of  the  nudibranchs  gave  them  a  slight  iridesence  in  contrast  to  the  hy- 
droid, so  that  even  very  small  nudibranchs  (<0.5  mm)  could  be  seen  among  the 
polyps. 

Egg  masses  laid  in  the  laboratory  were  isolated  in  small  dishes  containing  50  ml 
of  sea  water  and  incubated  at  11-13°  C.  The  sea  water  used  for  culturing  was 
collected  in  Gosport  Harbor  and  filtered  through  a  0.45  /mi  Millipore  filter.  The 
culture  water  was  initially  changed  daily,  but  in  later  experiments  it  was  changed 
every  two  or  three  days  with  no  effect  on  development.  At  intervals  of  six  to 
twenty-four  hours,  the  egg  masses  were  temporarily  transferred  in  drops  of  sea 
water  to  microscope  slides  and  observed  under  a  compound  microscope  using  trans- 
mitted or  reflected  illumination.  Egg  masses  collected  in  the  field  were  cultured  at 
the  temperature  at  which  they  were  collected,  which  ranged  from  4-13°  C. 

The  normal  mode  of  development  for  Cuthona  nana  eggs  cultured  at  11-13°  C 
was  determined  initially,  then  the  effect  of  variations  in  temperature,  adult  nutri- 
tion, and  the  presence  of  H.  echinata  was  tested.  Specimens  of  C.  nana  and  H. 
echinata  colonies  were  kept  in  dishes  of  aerated  sea  water  at  4,  8,  or  16°  C.  De- 
posited egg  masses  were  isolated  as  above  and  incubated  at  the  same  three  tempera- 
tures. At  16°  C,  successful  development  was  obtained  only  when  the  water  was 
changed  at  least  twice  daily.  Egg  masses  from  starved  adults  were  isolated  at  1 1- 
13°  C  and  their  development  followed. 

Other  specimens  of  Cuthona  nana  were  kept  in  compartmentalized  trays  with 
flowing  sea  water.  Shells  covered  with  H .  echinata  were  included  with  some  speci- 
mens of  C.  nana.  The  growth  of  individual  nudibranchs  could  thus  be  followed,  the 
availability  of  food  controlled,  and  the  number  of  egg  masses  laid  by  particular 
individuals  monitored. 

Although  the  behavior  and  distribution  of  the  early  postlarval  stage  of  C.  nana 
could  not  be  directly  studied  in  the  field,  the  distribution  of  juveniles  on  H. 
echinata  colonies  was  noted  monthly  and  two  field  experiments  wrere  conducted  to 
test  for  the  presence  of  planktonic  C.  nana  veligers.  In  the  first,  a  float  was 
anchored  3  m  off  the  bottom  of  Haley's  Cove  in  8  m  of  water  on  February  24,  1974. 
Pairs  of  H.  cchina to-covered  shells  were  suspended  at  1,  2,  and  3  m  off  the  bottom 
to  determine  if  C.  nana  veligers,  should  they  be  capable  of  swimming,  would  settle 
directly  on  an  H.  echinata  colony.  These  hydroid  colonies  had  been  collected  ap- 
proximately two  weeks  earlier,  and  had  been  examined  under  the  dissecting  micro- 
scope to  remove  all  specimens  of  C.  nana.  The  second  experiment  involved  anchor- 
ing a  1  X  1  X  0.5  m  open-bottomed  cage  covered  with  0.25  inch  nylon  mesh  on  the 
sand  near  the  float.  The  hermit  crabs  initially  enclosed  by  the  cage  were  removed 
before  nine  hermit  crabs  bearing  H.  echinata-covered  shells  were  placed  inside. 
These  colonies  had  also  been  cleaned  of  C.  nana.  It  was  hypothesized  that  plank- 
tonic  C.  nana  veligers  might  settle  near  H.  echinata  colonies  before  metamorphosing 


DEVELOPMENT  OF  CUTHONA  NANA 


159 


FIGURE  1.  Two  specimens  of  the  eolid  nudibranch  Cuthona  nana  feeding  on  the  colonial 
hydroid  Hydractiiiia  echinata  covering  a  gastropod  shell  occupied  by  the  hermit  crab  Pagurus 
acadianus.  The  larger  nudibranch  is  about  14  mm  in  length. 

and  climbing  onto  the  hydroid.  The  shells  suspended  from  the  float  or  confined 
inside  the  cage  were  changed  at  two  week  intervals  until  May  29,  1974.  Each 
time,  the  hydroid  colonies  were  examined  for  nudibranchs  immediately  upon  re- 
turn to  the  laboratory,  and  again  one  and  two  weeks  later. 

RESULTS 
Development 

Egg  mass.  Cuthona  nana  is  reproductively  typical  of  opisthobranchs  in  that  it  is 
a  reciprocally  copulating  hermaphrodite  that  deposits  eggs  within  a  gelatinous 
stroma.  The  spawn  of  C.  nana  has  been  described  and  illustrated  by  Harris  et  al. 
(1975).  The  largest  egg  masses  collected  in  the  field  or  laid  in  the  laboratory  were 
10  mm  in  diameter  and  contained  about  1500  eggs.  However,  most  egg  masses 
observed  were  considerably  smaller,  averaging  450  eggs.  Nudibranchs  raised  in 
the  laboratory  from  immaturity  to  death  laid  up  to  16  egg  masses.  The  first  and 
last  few  egg  masses  laid  were  smaller  than  average,  but  most  contained  300-600 
eggs.  Individuals  separated  after  copulation  laid  up  to  six  egg  masses  before  a 
substantial  number  of  unfertilized  eggs  were  produced. 


100  P.KI  AX    R.    RIYIiST 

TAHU:   I 
Normal  table  of  development  for  Cuthona  nana  eggs  incubated  ut  II    13°C. 


\  2  hours 

First  division 

6    10  davs 

Cilia  and  shell  develop 

16  hours 

Second  division 

15  davs 

Mantle  withdrawn  half-w 

ay 

20  hours 

Third  division 

16  days 

Propodium  lirst  appears 

36  hours 

Morula 

18-21  days 

Hatching 

3-5  days 

Gastrulation 

20-23  days 

Metamorphosis 

Development  to  hatching.  Karly  development  in  Cuthona  nana  is  similar  to 
that  described  for  other  opisthobranchs  (Casteel,  1904;  Pelseneer,  1(H1;  Thompson, 
1958).  Table  I  gives  normal  development  time  for  eggs  cultured  at  11-13°  C.  At 
oviposition,  tbe  white  eggs  within  their  individual  ovate  capsules  average  160  //.m 
in  diameter.  Spiral  cleavage  produces  a  stereoblastula  whose  vegetal  side  begins  to 
flatten  at  the  end  of  the  second  day  of  development.  Gastrulation  results  in  a  cup- 
shaped  gastrtila,  with  the  ventral  blastopore  becoming  asymmetrical  before  closing 
during  the  fifth  day.  Typically,  the  polar  bodies  adhere  to  the  animal  pole  through 
gastrulation. 

liv  the  end  of  the  sixth  day  a  shell  cap  covers  the  posterior  end  of  the  embryo 
(Fig.  2a).  The  shell  increases  in  size  as  the  shell  gland  (now  the  mantle  fold) 
spreads  anteriorly.  Two  anal  cells  of  disputed  function  (see  Bonar  and  Hadfield, 
1974)  appear  in  front  of  the  mantle  fold  on  the  right,  ventro-lateral  surface  of  the 
embryo,  while  anteriorly  the  velar  lobes  and  foot  are  enlarging.  The  locomotor 
cilia  elongate  and  rock  the  embryo  within  the  egg  capsule.  The  rate  of  shell  forma- 
tion exceeds  the  speed  at  which  the  mantle  fold  migrates  anteriorly,  so  that  a  lumen 
(the  perivisceral  cavity)  develops  in  the  posterior  end  of  the  shell  (Fig.  2b-c  ).  The 
visceral  mass  is  compact  and  opaque,  occluding  the  anterior  opening  of  the  develop- 
ing shell.  The  retractor  muscle  is  visible  within  the  perivisceral  cavity,  but  shows 
no  signs  of  contracting  during  shell  formation.  A  group  of  cells  surround  its 
origin  just  dorsal  and  to  the  left  of  the  shell  apex.  The  anal  cells  remain  visible 
for  a  time  in  front  of  the  mantle  fold  but  disappear  before  the  shell  becomes  com- 
plete on  the  tenth  day.  Torsion  in  C.  nana  does  not  involve  a  180°  twist  of  the 
cephalo-pedal  elements  with  respect  to  the  shell;  these  parts  differentiate  in  their 
post-torsional  positions.  The  movement  of  the  anal  cells  may  be  the  only  onto- 
genetic  evidence  of  torsion  (Thompson,  1962). 

During  shell  growth  the  foot  elongates  ventrally  and  becomes  heavily  ciliated 
mid-ventrally  and  at  the  tip,  but  not  laterally.  Its  dorsal  surface  has  an  operculum 
by  the  ninth  day  and  several  long,  stiff  compound  flagella  protrude  from  the  tip. 
At  this  time,  a  ciliated  subvelar  ridge  begins  to  develop.  The  visceral  organs  grow 
posteriorly  into  the  perivisceral  cavity,  eventually  filling  the  entire  posterior  end  of 
the  shell  except  for  the  dorsal  mantle  cavity.  Due  to  the  volk  content  and  opacity, 
it  is  difficult  to  discern  individual  organs. 

When  secretion  of  the  larval  shell  is  complete,  the  mantle  fold  at  the  shell 
aperture  becomes  thinner  and  less  dense.  By  the  eleventh  day,  it  begins  to  with- 
draw posteriorly  along  the  inner  surface  of  the  shell.  The  degree  to  which  the 
velar  lobes  can  be  retracted  into  the  mantle  cavity  depends  on  the  position  of  the 
withdrawing  mantle  fold.  Initially,  the  velum  cannot  be  accommodated  by  the 


DEVELOPMENT  OF  CUTHONA  NANA 


161 


MF 


AC 


VM 


VM 


VM 


FIGURE  2.  Late  larval  development  and  early  metamorphosis  in  Cnthoiw  innia:  a-c )  veli- 
ger  with  developing  shell;  d)  veliger  with  complete  shell  and  mantle  withdrawn  back  alon.u 
inside  of  the  shell;  e)  late  veliger  retracted  into  the  shell  at  the  stage  of  development  at 
hatching ;  and  f )  early  metamorphosis  after  loss  of  the  velum.  AC  indicates  the  anal  cells  ;  E, 
the  eye;  F,  the  metapodium  ;  MF,  the  mantle  fold  ;  OP.  the  operculum ;  PR,  the  propodium  ; 
RM,  the  retractor  muscle;  SH,  the  shell;  ST,  the  statocyst ;  SV,  the  suhvelar  ridge;  V,  the 
velum;  and  YM,  the  visceral  mass.  Yelar  locomotor  cilia  not  shown. 

mantle  cavity,  and  only  when  the  mantle  has  regressed  three  quarters  of  the  way  to 
the  apeyi  of  the  shell  (Fig.  2d)  can  the  velar  lobes  be  entirely  withdrawn.  Normally, 
the  foot  is  never  fully  retracted  with  the  operculum  closing  the  shell  opening. 
Chemical  irritants  such  as  alcohol  canst  the  veliger  to  retract  beyond  the  normal 
restrictions  of  the  mantle  fold  or  even  to  draw  the  foot  into  the  shell,  but  only  at 
concentrations  that  kill  the  larva. 

Two  red  eyespots  become  visible  on  the  fifteenth  day.  A  day  later,  the  pro- 
podium  begins  to  form,  just  ventral  to  the  mouth  (Fig.  2d).  At  this  stage  the 
mantle  has  migrated  two-thirds  of  the  way  back  from  the  shell  aperture.  The  sub- 


162 


BRIAN  R.  RIVEST 


OP 


,Y  ST 


STO 


100 


DG 


FIGURE  3.  Late  metamorphosis  and  early  postlarval  development  in  Cuthona  naiia:  a) 
shell  loss;  b)  newly  emerged  juvenile;  c)  elongated  juvenile;  and  d)  juvenile  with  four 
primary  cerata.  Stages  a-c  occur  in  20-23  days  after  oviposition  at  11-12°  C.  Growth  to  stage 
d  occurs  in  another  2-3  weeks  in  the  presence  of  abundant  food.  BM  indicates  the  buccal  mass; 
C,  two  cerata;  DG,  the  digestive  gland;  E,  the  eye;  OP,  the  operculum ;  SH,  the  shell;  STO, 
the  stomach ;  VM,  the  visceral  mass ;  and  Y,  yolk  concentrations  in  the  visceral  mass. 


velar  ridge  is  well-developed  and  heavily  ciliated.  By  the  eighteenth  day  the  mantle 
has  reached  the  shell  apex,  and  fuses  with  the  epithelial  layer  covering  the  visceral 
mass.  The  veliger  is  attached  to  the  shell  only  around  the  origin  of  the  larval 
retractor  muscle,  and  possibly  ventrally.  In  another  day  the  propodium  is  fully 
developed  (Fig.  2e),  and  muscular  activity  in  the  foot  is  evident.  The  velar  locomo- 
tor  cilia  continue  to  beat  almost  continuously,  but  the  veligers  move  about  very 
little  within  the  capsules.  The  visceral  organs  are  still  densely  packed  with  yolk, 
while  the  foot  and  velum  have  become  progressively  less  opaque. 

Hatching.  The  rate  of  development  varies  slightly  among  siblings,  although 
there  is  no  noticeable  relationship  between  position  in  the  egg  mass  and  develop- 
mental rate.  The  hatching  of  larvae  from  an  egg  mass  is  usually  spread  over 
several  days.  At  11-13°  C,  a  few  veligers  typically  begin  escaping  from  their 
capsules  late  on  the  eighteenth  day  after  oviposition.  Cracks  develop  and  radiate 
throughout  the  egg  capsule  causing  its  collapse.  Although  at  this  time  the  uniseriate 


DEVELOPMENT  OF  CUTHOXA  XAXA  163 

radula  possesses  two  distinct  teeth,  it  is  not  used  to  rupture  the  capsule  wall. 
Furthermore,  the  physical  activity  of  the  veligers  does  not  change  markedly  prim- 
to  hatching.  It  is  thus  unclear  what  causes  the  breakdown  of  the  capsules.  Also, 
hatching  from  the  capsules  does  not  depend  on  the  integrity  of  the  egg  mass. 
Usually  the  gelatinous  stroma  of  the  egg  mass  deteriorates  before  the  eggs  hatch, 
but  when  incubated  in  still,  filtered  sea  water  this  outer  covering  often  remains 
intact.  The  process  of  hatching  is  identical  in  either  situation. 

The  behavior  of  newly  hatched  veligers  depends  on  their  relative  stage  of 
development.  Yeligers  may  hatch  prior  to  the  complete  development  of  the  pro- 
podium.  These  larvae  initially  lay  on  their  sides  with  the  velum  extended  and 
velar  cilia  beating.  Although  the  cilia  may  gently  rock  the  veliger,  or  even  lift  the 
anterior  end  off  the  substratum  so  that  it  faces  upward,  a  veliger  was  never  seen 
to  swim  up  into  the  water.  When  the  propodium  is  more  fully  developed,  the 
veligers  roll  over  and  crawl  slowly  about,  with  the  velum  only  partially  extended. 
Veligers  that  hatch  with  a  well-developed  propodium  immediately  adhere  to  the 
bottom  and  begin  crawling.  Movement  may  continue  for  up  to  two  days,  but 
activity  progressively  decreases. 

MetanwrpJiosis.  Within  a  day  or  two  of  hatching,  metamorphosis  begins.  The 
veligers  cease  crawling  and  remain  in  a  semi-contracted  state,  with  the  shell  held 
nearly  vertically  and  the  head  just  inside  the  shell  aperture.  The  first  noticeable 
morphological  change  is  the  loss  of  the  velar  lobes.  The  beat  of  the  locomotory 
cilia  becomes  increasingly  erratic.  Cells  bearing  these  cilia  are  cast  off  from  the 
velum.  In  individuals  still  held  within  the  egg  mass  stroma.  these  cells  do  not 
accumulate  but  appear  to  be  ingested,  as  are  homologous  cells  shed  during  meta- 
morphosis in  Phcstilla  sibogac  (Bonar  and  Hadfield,  1974).  The  rest  of  the  velar 
lobe  is  resorbed  during  the  next  several  hours,  until  only  a  swelling  remains  pro- 
truding slightly  from  the  dorsum. 

During  this  period,  loss  of  contact  between  the  larval  body  and  shell  continues 
ventrally  until  only  the  retractor  muscle  attachment  remains  (Fig.  2f).  The  oper- 
culum  is  still  attached  to  the  dorsal  surface  of  the  metapodium  but  becomes  pro- 
gressively detached  distally,  allowing  increased  flexibility  of  the  foot.  Activity  of 
the  retractor  muscle  diminishes.  By  the  time  the  velar  lobes  are  resorbed,  me- 
chanical and  nonlethal  chemical  stimuli  do  not  elicit  further  withdrawal  of  the 
larva  into  the  shell.  Should  the  larva  become  dislodged  from  the  substrate,  the 
pedal  cilia  slowly  spin  it  about  until  a  foothold  is  regained. 

Shell  loss  begins  when  a  strong,  continuous  contraction  of  the  retractor  muscle 
breaks  its  connection  with  the  shell.  This  occurs  only  if  the  foot  is  firmly  attached 
to  the  substratum.  For  several  hours  after  the  connection  is  broken,  the  remnant 
of  the  origin  of  the  retractor  muscle  may  be  visible  as  a  small  lump  of  cells  or 
thickening  in  the  epidermal  layer.  In  some  metamorphosing  larvae  this  lump  is 
not  seen,  possibly  due  to  a  more  contracted  state  of  the  retractor  muscle  after  it 
detaches  from  the  shell.  With  the  retractor  muscle  attachment  severed,  the  larvae 
are  free  to  crawl  out  of  the  shell  (Fig.  3a),  a  process  that  may  take  five  to  eight 
hours  in  still  water.  Loss  of  the  shell  may  be  greatly  accelerated  by  water  cur- 
rents, because  the  visceral  mass  is  compact  and  equal  to  or  just  smaller  in  diameter 
than  the  shell  aperture.  The  operculum  may  adhere  to  the  shell  when  it  is  cast 
off  or  is  lost  separately. 


164  BRIAN  R.  RIVEST 

The  visceral  mass  of  the  young  shell-less  juvenile  initially  appears  as  a  distinct 
hump  (Fig.  3b).  Within  a  day  it  fuses  with  the  cephalo-pedal  elements,  thereby 
flattening  the  nudibranch  dorso-ventrally.  Elongation  of  the  body  continues  until 
the  juvenile  measures  about  0.32  mm  in  length  (Fig.  3c).  The  developing  buccal 
mass  possesses  three  to  four  teeth  in  the  imiseriate  radula  and  a  pair  of  weak  jaws. 
The  body  is  generally  a  translucent  white  color,  with  the  red  eyes  clearly  visible 
near  the  cerebral  ganglia.  The  visceral  mass  is  cream-white,  indicating  that  yolk 
still  remains.  Cilia  densely  cover  the  ventral  epithelium,  but  are  sparse  on  the 
dorsum. 

The  effect  of  different  temperatures  on  development.  The  development  of  eggs 
laid  and  maintained  at  4,  8,  or  16°  C  differed  from  those  kept  at  11-13°  C  only  in 
the  length  of  time  until  metamorphosis.  Egg  size  did  not  vary  with  temperature. 
Metamorphosis  occurred  within  50-55  days  at  4°  C,  34-36  days  at  8°  C,  and  16-17 
days  at  16°  C.  Cuthona  nana  may  not  tolerate  temperatures  much  above  16°  C, 
for  specimens  maintained  at  that  temperature  suffered  a  high  rate  of  mortality. 
Eight  adults  placed  with  food  at  16°  C  died  within  10  days.  Only  5-30%  of  the 
eggs  from  spawn  laid  at  16°  C  or  transferred  to  that  temperature  from  11-13°  C 
immediately  following  oviposition  developed  normally  through  metamorphosis.  In 
contrast,  nearly  all  of  the  eggs  laid  and  maintained  at  11-13°  C  reached  metamor- 
phosis. Nine  egg  masses  collected  in  the  field  and  incubated  at  the  temperature  at 
which  they  were  collected  (4-13°  C)  developed  and  metamorphosed  normally, 
with  their  rates  of  development  varying  according  to  the  temperature,  and  they  are 
identical  with  those  observed  for  eggs  raised  in  the  lab  at  similar  temperatures. 

The  effect  of  adult  nutrition  and  the  presence  of  Hydractinia  echinata  on  larval 
development.  Hatching  and  the  events  of  metamorphosis  proceeded  sequentially, 
unaffected  by  the  presence  of  H.  echinata  during  any  stage  of  development.  The 
development  of  embryos  in  egg  masses  incubated  in  a  dish  with  H.  echinata  and 
those  exposed  to  the  hydroid  only  before  or  after  hatching  did  not  differ  from  the 
development  of  those  egg  masses  kept  isolated.  The  integrity  of  the  spawn  mass 
had  no  effect.  The  development  and  metamorphosis  of  larvae  from  egg  masses  that 
had  broken  down,  exposing  the  capsules  directly  to  the  water,  was  similar  to  those 
in  egg  masses  that  remained  intact. 

Adult  specimens  of  C.  nana  kept  without  food  in  Millipore-filtered  sea  water 
continued  to  lay  egg  masses  for  up  to  ten  days.  The  size  and  activity  of  the  adults 
progressively  diminished  during  that  time,  but  some  starved  individuals  survived  for 
17  days.  They  laid  several  egg  masses  in  the  first  few  clays  of  isolation,  but  the 
frequency  of  oviposition  and  the  number  of  eggs  per  spawn  decreased  with  time. 
Mean  egg  diameter  did  not  vary,  and  the  rate  and  events  of  larval  development  and 
metamorphosis  proceeded  normally  in  the  presence  or  absence  of  H.  echinata. 

Postlarval  development  to  the  adult  stage.  In  the  absence  of  H.  echinata,  the 
newly  metamorphosed  juveniles  crawl  about  constantly.  They  are  negatively  geo- 
tactic  and  positively  phototactic,  crawling  up  the  sides  of  the  dishes  and  toward 
unidirectional  illumination  or  to  the  apex  of  rocks  or  shell  fragments  placed  in  their 
dishes.  When  H.  echinata  is  added,  movement  is  directed  toward  the  hydroid. 
However,  in  the  absence  of  food,  activity  decreases.  Starved  postlarvae  become 
motionless  within  two  weeks  of  metamorphosis,  but  movement  increases  rapidly 
when  //.  echinata  or  water  exposed  to  the  hydroid  is  added.  Postlarvae  survived 


DEVELOPMENT  OF  CUTHONA  NANA  165 

for  five  to  six  weeks  without  food  at  11-13°  C  and  ten  weeks  at  4°  C,  and  were 
then  still  capahle  of  feeding  and  growing  if  H.  cdiinata  was  made  available.  Post- 
larvae  from  four  egg  masses  laid  by  starved  adults  at  11-13°  C  survived  up  to  six 
weeks  without  food.  As  the  juveniles  are  starved,  the  visceral  mass  becomes  less 
intensely  colored.  Teeth  are  added  to  the  radula  until  there  are  five  to  eight,  but 
no  more  are  formed  unless  feeding  begins. 

Cuthona  nana  is  attracted  to  H.  cchinata  almost  immediately  after  metamor- 
phosis, but  feeding  is  initially  very  slow  if  it  occurs  at  all.  In  animals  that  crawl 
onto  H.  cchinata  shortly  after  metamorphosis,  the  orange  color  of  the  hydroid  does 
not  appear  in  the  digestive  gland  of  the  nudibranchs  for  two  or  three  days.  How- 
ever, juveniles  that  are  not  given  food  until  four  or  five  days  following  shell  loss 
begin  feeding  immediately  and  color  appears  in  the  digestive  gland  within  24  hours. 
Development  of  the  postlarval  buccal  or  digestive  structures  necessary  for  feeding 
may  therefore  continue  for  several  days  following  shell  loss. 

Cuthona  nana  will  readily  feed  on  any  part  of  H.  cchinata  colonies,  as  well  as 
released  eggs,  planula  larvae,  or  metamorphosing  planulae.  Young  nudibranchs  are 
sometimes  much  smaller  than  the  hydranth  they  are  feeding  on,  especially  the 
gastrozooids.  These  polyps  are  very  distensible  and  may  reach  a  length  of  5  mm  or 
more  and  an  oral  disc  diameter  of  more  than  0.75  mm.  Many  of  the  prey  items 
ingested  by  the  gastrozooids  are  much  larger  than  the  newly  metamorphosed  nudi- 
branchs, but  recently  collected  and  presumably  well-fed  colonies  of  H.  ecJiinata  do 
not  attempt  to  ingest  the  small  eolids.  On  such  colonies  maintained  in  still  water, 
postlarvae  are  often  seen  on  the  manubrium  of  gastrozooids.  Small  nudibranchs 
sometimes  elicit  a  defensive  response  by  H.  cdiinata;  the  tentacles  of  several  nearby 
polyps  are  brought  down  on  top  of  them,  but  the  hydroid's  nematocysts  apparently 
do  no  harm.  In  contrast,  hydroid  colonies  starved  for  several  weeks  will  eat 
recently  metamorphosed  C.  nana  juveniles.  The  nudibranchs  are  not  killed  before 
ingestion,  and  will  survive  if  immediately  removed  from  the  gastrocoel  of  the  hy- 
dranth. Those  removed  an  hour  or  more  later  are  dead  and  partly  digested. 

Table  II  summarizes  post-metamorphic  growth  in  C.  nana.  Growth  is  initially 
slow,  even  with  an  abundance  of  H.  cchinata.  Dorsal  enlargements,  indicating  the 
rudiments  of  the  first  pair  of  cerata,  do  not  appear  for  two  weeks  after  metamor- 
phosis. The  second  pair  of  ceratal  buds  develop  posterior  to  the  first  pair  within 
another  five  days  (Fig.  3d).  New  cerata  develop  at  an  increasingly  rapid  rate, 
with  their  pattern  of  appearance  like  that  described  for  Cuthona  adyarcnsis  by  Rao 
(1961).  The  uniseriate  radula  also  grows  in  size,  with  new  teeth  being  distinctly 
larger  than  the  first  five  or  six.  Eventually  C.  nana  adults  may  attain  a  length  of 
28  mm  with  250  cerata  and  27  radular  teeth,  but  nudibranchs  measuring  17-22 
mm  with  23-24  radular  teeth  are  more  common. 

Development  and  growth  varies  substantially  in  individuals  from  the  same  egg 
mass,  so  that  some  mature  several  weeks  before  others.  The  white  ovotestis  first 
becomes  visible  through  the  body  wall  when  the  nudibranchs  are  8-10  mm  long. 
Anterior  acini  of  the  ovotestis  develop  first,  and  maturation  proceeds  posteriorly. 
Nudibranchs  smaller  than  10  mm  have  never  been  seen  to  copulate,  and  they  are 
usually  longer  than  12  mm  before  they  lay  eggs.  In  the  laboratory,  specimens  of 
C.  nana  raised  on  healthy  H.  ediinata  on  a  gastropod  shell  remained  on  that  colony 
until  they  were  at  least  10  mm  long,  regardless  of  the  presence  or  absence  of  a 


166  BRIAN   K.    KIYKST 

TAm.ii   II 

Postldrnil  growth  in  ("utliona  nana  /;/  the  presence  of  abundant  Hydractinia  cchinata 

at  11-13°C. 


Time  from 
metamorphosis 

Length 
in  mm 

Characteristics 

2  weeks 

0.5 

Eight  to  ten  radular  teetli  ;  iirst  pair  of  cerata 

3  weeks 

0.75 

Eleven  to  twelve  radular  teeth;  third  pair 

of  cerata; 

rhinophore  primordia 

4^5  weeks 

1.0 

Fifteen  radular  teeth  ;  fourteen  cerata  ;  first 

heart  beat  ; 

oral  tentacles  appear 

6—7  weeks 

4.0 

Thirty  to  thirty-five  cerata 

9-10  weeks 

8.0 

Nineteen  to  twenty-four  radular  teeth 

1  1  weeks 

12.0 

Shortest  time  observed  for  an  individual 

to  mature 

and  lay  an  egg  mass 

hermit  crab  in  the  shell.  At  this  time,  if  alone,  C.  nana  leaves  the  hydroid  colony 
in  search  of  a  mate.  \Yhen  a  nudibranch  measuring  only  10-12  mm  mates  for  the 
first  time,  it  usually  resumes  feeding  before  laying  egg  masses. 

Harris  ct  al.  (1975)  reported  that  Cutliona  nana  adults  do  not  lay  egg  masses 
on  H.  cchinata  colonies.  They  noted  that  since  the  nudibranchs  do  not  die  after 
laying  eggs,  they  probably  find  new  hydroid  colonies.  Subsequent  laboratory 
observations  were  made  on  C.  nana  confined  with  hermit  crabs  bearing  PI.  cchinata- 
colonized  shells.  The  nudibranchs  invariably  left  the  hydroid  to  lay  egg  masses  and 
consistently  returned  to  the  colonies  to  resume  feeding.  The  hermit  crabs  often  re- 
mained motionless  long  enough  for  the  nudibranch  to  find  and  crawl  onto  the  hy- 
droid. This  pattern  of  leaving  the  //.  cchinata  to  deposit  spawn,  and  then  return- 
ing, continued  until  the  nudibranchs  died.  At  11-13°  C,  adult  specimens  of  C.  nana 
survived  in  this  fashion  for  six  to  eight  weeks,  while  those  kept  at  4°  C  lived  for 
over  three  months. 


•  300 

_                                                                                                                                                                      — 

18 

Ul 

m                                                                                                               • 

0) 

/•- 

16 

H 

_c 

.                                  N                                                                                                                                                                                                                                                 /          ' 

w  250 

'              •*                                                                /   ! 

_ 

o 

/     /   ~ 

14 

•c 

o 

/      ' 

re 

\200 

-                          \                                                 /        / 

,_    /       ;    - 

12 

01 

ra 

/      • 

c 

c 

10 

/                '     - 
'                                                      /                 ' 

10 

m 

c    150 

-.       / 

_. 

(0 

.^•/                                   \         / 

8 

3 

c 

O 

o 
£    100 

_•-             ^                               x.  /                 ^:' 

6 

O 

** 

•^                                       /*~*             ^"^ 

H 

3 

O 

°x"'                                                                                                                                             /            ^^ 

••-      °      ^m                                                                                            /                          -• 

4 

• 

-      50 

-   ""•'*                              *\          / 

o 

xv    / 

2 

o 

\  /     i 

i     i     i     i     i     i     i     i     i     i  •  i          i     i     i     i     i 

J         FMAMJJASONDJ         FMAMJ 

1974                                                                                                         1975 

FIGURE  4.  Number  of  Cuthona  uana  per  100  Ilydractin'm  cchiimla  colonies  and  surface 
water  temperatures  from  January,  1974,  to  June  1975.  Open  circles  refer  to  collections  outside 
of  Haley's  Cove.  No  collections  were  made  in  June,  July  and  August,  1974.  See  text  for  details. 


DEVELOPMENT  OF  CUTHOXA  NANA  167 

Two  generations  of  Cnthona  nana  were  raised  in  the  laboratory  at  11-13°  C. 
The  midibranchs  survived  well  in  the  recirculating  seawater  systems  where  the 
salinity  varied  from  30  to  38^,.  The  shortest  life  cycle  observed  took  14  weeks 
from  egg  to  egg. 

Ecology 

Citfhona  liana  was  found  to  be  more  abundant  in  Gosport  Harbor  than  pre- 
viously reported  (Harris  ct  a!.,  1975).     Both  juvenile  and  adult  individuals  were 
most  common  from   |anuarv  through  June  and  least  abundant  in  September  through 
December  (Table  III  ).     The  number  of  C.  nana  observed  in  February  and  March, 
1974,    were   considerably   below   that   seen    in   those   months   during    1975.      Foul 
weather  in  February  and  March,   1974,  forced  collection  outside  of  Haley's  Cove, 
the  area  where  the  density  of  pagurid  crabs  with  Hydractinia  cchinata  colonies  was 
highest  in  Gosport  Harbor.     Laboratory  studies  have  shown  that  the  growth  rate 
of  C.  uana  at  water  temperatures  observed  during  these  months   (3.5-4.5°   C)   is 
extremely  slow.     Therefore,  the  number  of  nudibranchs  found  in  April  and  May, 
1974,  when  the  temperature  reached  only  6.5°  C,  indicates  that  nudibranch  densi- 
ties within  Haley's  Cove  during  the  previous  two  months  were  much  higher  than 
in  the  collections  outside   Haley's   Cove.     Nudibranch   numbers   fluctuated   asyn- 
chronously  with  temperature  (Fig.  4).     The  population  of  C.  nana  was  growing  in 
size  during  the  coldest  months  and  decreased  sometime  during  the  summer.    The 
average  size  of  the  nudibranchs  collected  varied  little,  for  the  number  of  adults  and 
juveniles  varied  synchronously   (Table  III).     Egg  masses  were  seen  in  the  field 
during  every  month  of  this  study  except  for  September  and  November,  1974.     In 
October,  1974,  only  one  egg  mass  was  found,  and  only  two  were  found  in  December. 
Reflecting  the  greater  number  of  adults,  egg  masses  were  more  abundant  during 
the  spring  months  with  up  to  eight  seen  during  a  forty-minute  dive. 

Data  on  the  distribution  of  C.  nana  less  than  5  mm  in  length  are  included  in 
Table  III  to  show  monthly  differences  in  the  numbers  of  juveniles  and  to  evaluate 
seasonal  fluctuations  in  reproduction  and  recruitment.  Since  young  nudibranchs 
normally  do  not  leave  an  H.  cchinata  colon}-  until  they  are  about  8-10  mm  in 
length,  the  5  mm  length  was  considered  a  conservative  upper  size  limit  for  examin- 
ing the  distribution  of  nonreproductive  juveniles.  Such  juveniles  were  likely  to  be 
found  on  the  first  hydroid  colony  they  had  occupied. 

The  numbers  of  C.  nana  juveniles  fluctuated  from  a  spring  high  of  94  in  May, 
1974,  down  to  a  fall  low  of  7  in  Xovember,  1974,  and  up  to  242  per  100  hydroid 
colonies  in  June,  1975  (Table  III).  These  juveniles  were  not  evenly  distributed 
over  the  population  of  H.  ecJiinata  colonies.  In  the  months  when  most  abundant, 
young  nudibranchs  outnumbered  the  hydroid  colonies  collected,  yet  they  were  found 
on  only  48-57%  of  them.  Thus,  if  one  small  nudibranch  was  present  on  an  H. 
cchinata  colony,  chances  were  good  that  there  were  more.  Four  or  five  juveniles 
per  colony  were  not  uncommon,  and  much  higher  numbers  were  occasionally  found. 
In  May,  1974,  one  colony  collected  possessed  17  juveniles  less  than  3.5  mm  long.  A 
colony  examined  in  June  of  1975  carried  29  C.  nana  juveniles.  During  the  fall 
months  when  C.  nana  was  least  abundant,  H.  cchinata  colonies  with  two  or  more 
young  nudibranchs  were  still  more  frequently  encountered  than  those  with  just  one. 

Cuthona  nana  juveniles  were  also  not  randomly  distributed  over  the  H.  cchinata 


168 


BRIAN  R.  RIVEST 


TABLE  III 

Data  on  Cuthona  nana/r»w  monthly  collections  of  Hydractinia  echinata- 
covered  hermit  crab  shells  from  Gosport  Harbor. 


Month 

Number  of 
C.  nana  per 
100  shells* 

Number  of  C.  nana 
by  size  classes 

Average 
size  of 

C     Hfl  fit! 

Number 
of  hermit 
crab 

Percentage 

of  shells  with 
C.  nana 

Average 
number 
of  C.  nana 
<C5  mm  per 

<5 

Total 

<5 

5-10 

>10 

in  mm 

shells 
examined 

<5 

Total 

infested 
shell 

mm 

mm 

mm 

Jan.  1974 

90 

107 

26 

2 

3 

3.1 

29 

41 

59 

2.2 

Feb.** 

24 

67 

5 

8 

1 

6.1 

21 

14 

43 

1.6 

Mar.** 

9 

56 

2 

6 

4 

8.5 

22 

9 

36 

1.0 

Apr. 

89 

121 

34 

8 

4 

3.5 

38 

38 

42 

2.8 

May 

94 

138 

76 

19 

17 

4.5 

81 

44 

59 

2.1 

Sept. 

46 

51 

31 

4 

0 

2.6 

68 

24 

25 

1.9 

Oct. 

19 

31 

13 

6 

2 

4.0 

67 

10 

25 

1.8 

Nov. 

7 

11 

4 

1 

1 

4.4 

55 

7 

11 

1.0 

Dec. 

50 

53 

29 

1 

1 

1.8 

58 

31 

34 

1.6 

Jan.  1975 

128 

144 

32 

1 

3 

2.9 

25 

48 

56 

2.7 

Feb. 

123 

139 

54 

4 

3 

2.9 

44 

52 

61 

2.4 

Mar. 

152 

193 

70 

12 

7 

3.7 

46 

54 

72 

2.8 

Apr. 

143 

188 

80 

18 

7 

3.4 

56 

57 

71 

2.5 

May 

180 

238 

88 

15 

14 

2.8 

49 

55 

65 

2.7 

June 

242 

285 

138 

10 

15 

2.4 

57 

53 

70 

4.6 

X  =  2.7 

*  These  columns  were  obtained  by  adjusting  the  number  of  C.  nana  found  in  the  monthly 
samples  to  100  hermit  crab  shells  so  that  population  size  fluctuations  would  be  more  visible. 

**  Collections  during  these  months  were  made  outside  of  Haley's  Cove,  and  the  number  of  C. 
nana  found  was  lower  than  expected  for  Haley's  Cove.  See  text  for  details. 

colonies.  The  observed  distribution  of  nudibranchs  over  the  hydroid  colonies  col- 
lected each  month  was  compared  with  a  random  distribution  using  a  chi- 
square  analysis.  The  differences  were  significant  at  the  0.05  level  for  the  Haley's 
Cove  samples  for  all  months  except  October,  November,  and  December,  1974. 
Thus,  during  the  months  when  most  abundant,  the  juveniles  were  nonrandomly 
distributed  over  the  H.  cchinata  colonies. 

Two  observations  suggest  that  recruitment  of  C.  nana  on  hydroid  colonies  is 
from  benthic  juveniles  rather  that  pelagic  veligers.  First,  no  juveniles  were  found 
on  either  the  caged  pagurids  or  the  suspended  H.  cchinata  colonies  that  were  placed 
in  the  field  during  a  period  with  high  C.  nana  egg  production  and  a  growing  popula- 
tion (February  through  May,  1974).  In  the  laboratory,  starved  colonies  of  H. 
echinata  had  consumed  veligers  and  postlarvae  of  C.  nana.  Therefore,  the  hydroid 
colonies  recovered  from  the  float  and  cage  were  tested  to  determine  if  the  experi- 
mental manipulations  had  starved  them  to  a  point  where  they  might  have  eaten  any 
C.  nana  veligers  or  postlarvae  they  had  contacted.  Veligers  and  juvenile  nudi- 
branchs were  not  preyed  upon  when  placed  on  the  gastrozooids  of  experimental 
colonies  shortly  after  being  brought  back  from  the  field. 

Secondly,  C.  nana  juveniles  smaller  than  3  mm  in  length  were  found  pre- 
dominantly on  the  ventral  half  of  the  hydroid  colonies  collected.  The  gastrozooids 
of  H.  echinata  are  more  numerous  and  longer  around  the  ventral  periphery  of  the 


DEVELOPMENT  OF  CUTHONA  NANA  169 

colony.  These  polyps  sweep  over  the  surface  of  the  substrate  as  the  hermit  crab 
moves  about.  In  the  laboratory,  PL  ccliinata  colonies  swept  over  the  bottom  of 
dishes  containing  C.  nana  postlarvae  picked  up  many  of  the  small  nudibranchs  by  a 
mechanism  that  probably  involves  the  hydroid's  nematocysts.  These  collected 
nudibranchs  then  reoriented  and  began  feeding  on  the  hydranths.  By  the  time  they 
had  grown  to  a  length  of  5  mm.  they  may  have  moved  half  way  around  the  colony. 
Since  the  location  of  nudibranchs  less  than  3  mm  in  length  is  close  to  the  site  of 
first  contact  with  the  colony,  the  ventral  position  of  the  smallest  C.  nana  individuals 
on  the  collected  H.  ccliinata  colonies  supports  the  hypothesis  that  the  nudibranch 
reaches  the  hydroid  by  being  swept  up  from  the  bottom  and  not  by  settling  onto  the 
hydroid  from  the  plankton. 

In  the  field,  most  individuals  of  C.  nana  were  observed  on  //.  ccliinata  colonies, 
but  during  the  late  spring  months  adult  nudibranchs  were  often  found  crawling 
over  the  sand,  rubble,  or  loose  pieces  of  algae.  Occasionally  groups  of  two  to  four 
were  seen,  either  copulating  or  depositing  egg  masses,  but  most  were  isolated 
individuals.  Five  adult  nudibranchs  found  singly  on  the  bottom  or  on  a  hydroid 
colony  were  returned  to  the  laboratory  and  maintained  in  isolation.  In  every  case, 
fertile  egg  masses  were  subsequently  laid,  indicating  that  the  adults  had  copulated 
previously. 

In  late  April  and  May.  1974,  large  specimens  of  C.  nana  were  discovered  in  the 
cage.  During  each  of  four  two-week  periods,  four  to  six  nudibranchs  with  average 
lengths  of  16  mm  had  crawled  onto  the  hydroid  colonies,  and  several  more  were  seen 
in  and  around  the  cage.  The  presence  of  adults  of  C.  nana  on  the  caged  H.  ccliinata 
demonstrates  the  nudibranch's  mobility  and  capacity  for  finding  new  colonies. 

The  pagurid  population  in  Gosport  Harbor  consisted  of  Pagitrus  acadianus 
Benedict,  1901,  and  P.  arciiatiis  Squires,  1964.  Both  were  abundant  down  to  a 
depth  of  twelve  meters.  Pagnnts  acadianus  was  found  more  commonly  on  the 
cleaner  sand  and  P.  arciiatus  in  the  siltier,  more  cobble-strewn  areas,  but  there  was 
considerable  intermixing.  Observations  during  numerous  dives  indicated  that  the 
distribution  of  these  hermit  crabs  changed  continuously,  such  that  denser  concen- 
trations were  found  in  different  areas  of  Haley's  Cove  on  successive  dive  dates. 
Grant  (1963)  also  found  that  populations  of  P.  acadianus  in  the  shallow  subtidal 
were  transient  in  nature,  indicating  a  high  degree  of  mobility.  The  feeding  be- 
havior of  this  pagurid  has  not  been  described,  but  it  appears  to  be  similar  to  that 
of  the  omnivorous  European  species,  P.  bcrnhardus  (Orton,  1927;  Gerlach, 
Ekstro'm  and  Eckardt,  1976).  Food  of  P.  acadianus  consists  partly  of  moribund 
invertebrates  and  pieces  of  algae,  but  predominantly  of  detrital  material  and  small 
organisms  captured  by  using  the  chelae  to  shovel  sediment  into  the  mouth  parts 
where  it  is  sifted.  The  hermit  crabs  remained  stationary  much  of  the  time,  sifting 
sediment  or  actively  fanning  the  water  with  the  maxillipeds  and  maxillae,  possibly 
filter-feeding  as  in  P.  bcrnhardus  (Gerlach  ct  al,  1976).  Pagunis  acadianus  broods 
its  eggs  on  abdominal  pleopods  until  the  zoeal  stage.  Females  were  seen  in  ]\ larch. 
1974,  to  release  the  zoeae  by  protruding  three-fourths  of  the  way  out  of  their  shells 
and  waving  their  egg-laden  pleopods.  In  cases  where  the  shell  aperture  was  lined 
by  H.  cchinata,  some  zoeae  were  caught  and  eaten  by  the  gastrozooids.  Of  all  the 
hermit  crabs  collected  bearing  shells  colonized  by  H.  cchinata,  9S%  were  Pagnnts 
acadianus.  Like  the  European  P.  bcrnhardus  (Jensen,  1970),  P.  acadianus  prefers 


170  BRIAN  R.  RIVEST 

shells  covered  with  the  hydroid  over  clean  shells  (Grant  and  Ulnier,  1974).  In 
contrast,  P.  arena/its  preferentially  selects  naked  shells  (Grant  and  Ulnier,  1974). 
(Pagurus  f>u!>escens  in  Grant  and  Pontier,  1973,  and  Grant  and  Ulmer,  1974,  was 
actually  P.  arcuatus ;  personal  communication  from  W.  Grant,  1975.)  Whereas 
empty,  clean  gastropod  shells  were  commonly  seen  in  Gosport  Harbor  during  the 
present  study,  unoccupied  H .  echinata-covered  shells  were  rare. 

The  feeding  of  juvenile  nudihranchs  had  little  noticeable  effect  on  the  hydroid 
colonies;  the  regeneration  rate  of  the  hydranths  approximated  the  predation  rate. 
Adult  nudibranchs,  however,  cleared  patches  among  the  hydranths,  leaving  only  the 
basal  mat.  In  such  cases,  the  first  polyps  eaten  were  often  regenerating  while  the 
nudibranch  was  still  enlarging  the  patch. 

DISCUSSION 

The  pattern  of  development  of  Cuthona  nana  eggs  remained  constant  over  a 
variety  of  conditions  during  this  study.  Incubation  of  egg  masses  collected  in  the 
field  throughout  the  year  and  those  laid  in  the  laboratory  under  various  conditions 
of  temperature  and  adult  nutrition  yielded  veligers  which  invariably  metamorphosed 
without  a  pelagic  stage.  Developmental  rate  varied  inversely  with  temperature. 
The  times  to  hatching  obtained  at  4,  8,  11-13,  and  16°  C  fall  close  to  the  regression 
line  of  Spight  (1975,  Fig.  1)  for  prehatching  period  versus  temperature  for  other 
opisthobranchs.  These  results  further  support  his  thesis  that  time  to  hatching  can 
be  estimated  with  reasonable  accuracy  from  taxonomic  affinity  and  temperature 
alone. 

The  presence  or  absence  of  Hydractiriia  ccliinafa  had  no  noticeable  effect  on  the 
rate  or  sequence  of  events  in  the  development  of  C.  nana.  Egg  size  and  subsequent 
development  were  unaltered  by  differences  in  adult  nutrition ;  starved  animals 
simply  laid  fewer  eggs.  In  contrast  to  Mytilits  cdulis  (Bayne,  1972;  Bayne,  Gab- 
bott,  and  Widdows,  1975).  there  was  no  increase  in  abnormal  development  in  eggs 
from  starved  adults.  Furthermore,  starved  postlarvae  from  starved  adults  devel- 
oped as  fast  and  survived  as  long  (four  to  six  weeks  at  11-13°  C)  as  starved  post- 
larvae  produced  by  well-fed  adults.  Hydactinia  cchinata  apparently  plays  no  role 
in  inducing  metamorphosis  in  C.  nana,  as  Elect ra  pilosa  does  for  Aldaria  pro.vima 
(Thompson,  1958). 

Two  schemes  that  categorize  opisthobranch  development  have  been  presented  in 
the  literature.  Thompson  (1967)  formed  three  categories  distinguishing  opistho- 
branchs by  feeding  type  and  place  of  metamorphosis.  His  development-types  1,  2, 
and  3  refer  to  species  with  pelagic  planktotrophic,  pelagic  lecithotrophic  and  non- 
pelagic  lecithotrophic  ("direct")  development,  respectively.  Cuthona  nana  falls 
between  development-types  2  and  3  in  that  it  does  not  possess  a  pelagic  lecitho- 
trophic larva,  nor  does  it  hatch  out  of  the  capsule  at  a  post-veliger  benthic  stage. 
Because  of  the  ecological  significance  of  its  nonpelagic  development,  it  should  be 
classified  as  having  development-type  3.  In  this  category  Thompson  (1967)  in- 
cluded Cuthona  pustulata,  which  like  C.  nana  hatches  out  of  the  capsule  as  a  veliger, 
but  remains  within  the  stroma  of  the  egg  mass  until  metamorphosis  (Roginskaya, 
1962). 

Tardy  (1970)  presented  a  classification  scheme  for  the  Xudibranchia  that 
primarily  segregated,  them  on  the  basis  of  protoconch  type,  which  he  felt  represented 


DEVELOPMENT  OF  CUTHOXA  XANA  171 

basic  ontogenetic  differences  such  as  different  origins  of  the  adult  dorsal  epidermis. 
Species  with  a  spiral  protoconch  were  categorized  as  having  type  1  development, 
while  type  2  referred  to  those  species  possessing  an  inflated  protoconch.  There  are 
at  least  two  exceptions  to  Tardy's  scheme.  First,  Bonar  and  Hadfield  (1974)  and 
Bonar  (1976)  have  reported  that  the  dorsal  epidermis  of  Phestilla  sibogae  was 
derived  from  the  lateral  surfaces  of  the  larval  foot  and  not  from  the  floor  of  the 
mantle  cavity  as  thought  by  Tardy  for  type  2  nudibranchs.  Secondly,  whereas 
Tardy  felt  that  all  nudibranchs  with  inflated  protoconchs  underwent  torsion  after 
the  shell  was  complete,  in  Cittliona  nana  structures  develop  in  their  post-torsional 
positions.  Additional  studies  are  needed  on  the  origin  of  the  adult  dorsal  epidermis 
and  differences  in  the  expression  of  torsion  within  the  Opisthobranchia. 

The  field  data  support  the  laboratory  observations  of  nonpelagic  development  in 
Cuthona  nana.  A  pelagic  veliger  might  have  settled  on  the  experimental  colonies 
suspended  in  the  water  column  or  enclosed  by  the  cage,  but  this  was  not  observed. 
Table  III  shows  that  from  January  to  June,  1975,  juveniles  of  C.  nana  were  found 
on  only  about  one-half  of  the  hydroid  colonies  collected,  even  though  the  nudi- 
branchs greatly  outnumbered  the  colonies.  The  uneven  nonrandom  distribution 
observed  during  the  late  winter  and  spring  months  is  what  would  be  expected  if 
recruitment  to  the  C.  nana  population  was  from  simultaneous  colonization  by 
clustered  benthic  juveniles,  with  the  distribution  of  these  clusters  being  determined 
by  the  deposition  sites  of  egg  masses. 

Predatory  benthic  marine  invertebrates  that  are  relatively  nonmotile  as  adults 
and  lack  pelagic  larval  stages  are  faced  with  the  problems  of  prey  location  and  of 
dispersal.     From  field  and  laboratory  observations,  it  is  concluded  that  the  post- 
larvae  of  Cuthona  nana  'find'  an  Hydractinia  echinata  colony  much  the  same  way  as 
the  hydroid's  planulae  'find'  a  clean  hermit  crab  shell.     The  gonozooids  on  female 
hydroid  colonies  produce  large  orange-red  eggs  that  are  fertilized  when  released, 
drop  to  the  bottom  and  develop  within  two  or  three  days  into  a  planula  with  an  en- 
larged anterior  end  possessing  numerous  secretory  cells  (Bunting,   1894;  Van  de 
\yver,   1964,  1967).     A  healthy  colony  covering  an  hermit  crab  shell  may  release 
several  hundred  eggs  in  a  season.    The  resulting  planulae  remain  benthic  and  crawl 
slowly  about  in  a  turbellarian-like  fashion,  being  positively  phototactic  and  some- 
what negatively  geotactic  (Schijfsma.  1935;  Cazaux.  1961;  Van  de  Yyver,  1964). 
The  planulae  do  not  actively  search  for  a  clean  hermit  crab  shell;  it  is  the  hermit 
crab's  activity,  either  its  locomotion  or  feeding  movements,  that  bring  the  two  in 
contact   (Schijfsma,   1935).     The  planulae  adhere  to  the  shell  with  the  anterior 
end.    A  single  polyp  is  initially  formed,  then  a  basal  mat  grows  out  over  the  shell 
as  new  polyps  are  added.     The  gastrozooids  develop  primarily  around  the  ventral 
side  of  the  shell,  where  they  capture  small  invertebrates  on  the  surface  of  the  sub- 
strate during  the  hermit  crab's  travels  and  feeding  movements  (Christensen,  1967; 
Harris  ct  a!.,   1975).     Postlarvae  of  Cuthona   nana  also  get  picked  up  by   these 
gastrozooids.     Just  as  with  the  planulae,  the  positive  phototaxis  and  negative  geo- 
taxis  of  the  nudibranchs  keep  them  in  the  open  on  the  substrate  surface.     Such  posi- 
tioning increases  the  chances  that  the  postlarvae  will  be  swept  up  by  an  //.  echinata 
colony  should  an  hermit  crab  bring  one  by. 

The  mobility  of  the  hermit  crabs  is  likely  to  be  an  important  factor  in  the  spread 
of  Cuthona  nana  and  Hydractinia  echinata,  for  neither  species  has  an  actively  dis- 


172  BRIAN  R.  RIVEST 

parsing  larva.  An  adult  nudibranch  feeding  on  an  /-/.  ccliinata  colony  will  be  carried 
across  the  bottom  as  the  hermit  crab  wanders,  and  may  be  taken  tens  of  meters 
before  it  leaves  the  colony  to  find  a  mate  or  lay  eggs.  The  next  hydroid  colony  it 
climbs  onto  will  be  carried  in  a  direction  and  distance  independent  of  the  previous 
ones.  Large  Pagunis  acadianns  in  deeper  water  (below  17  in)  have  been  seen  in 
Lunatia  or  Bucciniun  shells  bearing  H.  cchinata  and  C.  nana.  These  large  crabs 
are  quite  noticeable  because  of  their  size,  but  are  relatively  rare.  Their  occasional 
presence  in  frequently  observed  areas  indicates  they  probably  travel  long  distances. 
They  are  sometimes  seen  in  shallow  water  where  small  hydroid-colonized  hermit 
crab  shells  are  more  numerous.  By  visiting  different  hermit  crab  concentrations, 
these  large  hermit  crabs  may  provide  a  means  for  colonizing  new  areas  and  a 
mechanism  of  genetic  communication  between  physically  distant  populations  of  C. 
nana.  Other  crab  species  may  also  be  important ;  C.  nana  was  collected  on  H. 
echinata  growing  on  the  legs  and  ventral  side  of  a  Cancer  borcalis  in  18  m  of  water. 
Hermit  crabs  also  act  as  colonization  vectors  for  the  direct  developing  Crepidula 
conve.va  (Hendler  and  Franz,  1971 ). 

Cuthona  nana  may  also  disperse  by  rafting  on  pieces  of  dislodged  algae.  Wave- 
dislodged  algae  occasionally  blanket  the  bottom  in  shallow  areas  of  Gosport  Harbor. 
Nudibranchs  that  climb  onto  the  algae  or  egg  masses  deposited  there  could  be 
carried  off  by  storms  or  current  changes.  Data  from  seabed  drifters  indicate  that 
the  local  average  water  speed  is  from  0.07  km/day  (Loder,  Anderson,  and  Shevenell, 
1973)  to  0.2  km/day  (Graham,  1970).  Thus  at  4°  C,  juveniles  having  developed 
and  metamorphosed  on  a  piece  of  drifting  algae  could  travel  8.75  to  25  km  before 
starving. 

The  interspecific  association  between  Pagurus  acadianns  and  Hydractinia  cchi- 
nata is  mutually  advantageous.  The  hermit  crab's  shell  provides  a  suitable  substrate 
for  the  hydroid,  which  in  turn  makes  the  shell  more  desirable  for  P.  acadianns  and 
less  so  for  P.  arciiatus  (Rees,  1967;  Grant  and  Pontier,  1973;  Grant  and  Ulmer, 
1974).  Pagurus  acadianns  will  even  occupy  shells  smaller  than  their  preferred  size 
range  if  these  shells  are  colonized  by  H.  cchinata  (Grant  and  Pontier,  1973).  The 
hydroid  can  increase  the  effective  size  of  the  shell  by  growing  beyond  the  lip,  so  the 
hermit  crab  needs  to  change  its  shell  less  frequently  (Harris  ct  al.,  1975;  Jensen, 
1975).  Hydractinia  cchinata  may  act  as  a  deterrent  to  predation  on  the  pagurid 
crabs.  Grant  and  Pontier  (1973)  found  that  Cancer  irroratus  did  not  feed  on  P. 
acadianns  occupying  shells  with  H.  cchinata  colonies.  However,  during  the  present 
study  a  few  starved  specimens  of  C.  borcalis  and  Carcinus  tnaenus  did  feed  on  P. 
acadianns  in  hydroid-covered  shells,  although  most  did  not. 

The  hermit  crab's  mobility  may  provide  a  means  of  escape  from  overpredation 
for  Hydractinia  cchinata.  The  hydroid  colonies  are  perennial  and  once  established, 
may  persist  for  years  (Sutherland,  1975;  personal  observation).  Cnthona  nana 
preys  on  the  hydroid  by  eating  the  polyps,  leaving  the  basal  mat  intact.  "When  an 
adult  nudibranch  leaves  an  hydroid  colony  to  search  for  a  mate  or  deposit  eggs,  the 
colony  will  be  carried  away,  descreasing  its  chances  of  being  preyed  upon  by  that 
nudibranch  again.  During  the  spring  and  early  summer  months,  when  large  speci- 
mens of  C.  nana  were  most  common,  the  majority  of  the  hydroid  colonies  collected 
possessed  patches  devoid  of  polyps  due  to  grazing  by  C.  nana,  the  only  significant 
local  predator  of  H.  cchinata.  However,  never  was  a  colony  collected  that  had 


DEVELOPMENT  OF  CUTHONA  NANA  173 

more  than  half  of  its  polyps  eaten  and  usually  the  grazed  areas  showed  signs  of 
regeneration.  Cuthona  nana  thus  appears  to  simply  crop  H.  cchinata  colonies  and 
not  kill  them,  with  the  perennial  colonies  regenerating  lost  polyps.  Similarly, 
Dendronofos  iris  feeds  on  just  a  few  tentacles  of  Cerianthns  sp.,  not  killing  the 
anemone  which  presumably  replaces  the  lost  tentacles  (Wobber,  1970). 

Prccuthona  pcacJii  (Alder  and  Hancock,  1847)  has  been  reported  to  feed  on 
Hydmctinia  growing  on  hermit  crab  shells  in  Europe  (Farran,  1903;  Swennen, 
1961  ;  Christensen,  1977).  Several  workers  ( L.  Harris,  T.  Gosliner,  T.  E.  Thomp- 
son, G.  Brown;  personal  communications)  consider  P.  peachi  to  be  a  junior 
synonym  of  Cuthona  nana.  Christensen  (1977)  recently  reported  that  P.  f>cachi 
in  Sweden  produced  actively  swimming  planktotrophic  veligers.  These  larvae 
survived  unfed  for  14  days  without  metamorphosing  in  the  presence  of  Hydractinia. 
Christensen  reported  an  egg  diameter  for  P.  peachi  of  100  /mi  and  a  development 
time  to  hatching  of  20-26  days  at  7-9°  C.  This  compares  with  the  respective 
values  determined  during  the  present  study  on  C.  nana  of  160  /mi  and  31-34  days 
at  8°  C,  with  the  veligers  remaining  benthic  and  metamorphosing  immediately  after 
hatching.  Different  modes  of  development  have  been  previously  reported  for  C. 
nana  by  Harris  ct  al.  (1975)  who  found  planktotrophic  development  in  1971  and 
nonpelagic  lecithotrophic  development  in  1973  in  egg  masses  laid  by  individuals 
collected  off  the  New  Hampshire  and  Maine  coasts.  These  differences  may  have 
resulted  from  observations  on  two  virtually  indistinguishable  species,  or  C.  nana 
may  indeed  possess  two  modes  of  development  with  the  factors  that  influence  the 
developmental  pattern  remaining  enigmatic. 


I  wish  to  thank  Larry  G.  Harris  for  his  stimulation  and  guidance,  Richard 
Strathmann  and  Alan  Kohn  for  critically  reading  early  drafts  of  the  manuscript,  and 
Alan  Kuzirian  for  his  friendly  assistance.  T.  E.  Thompson  and  Greg  Brown 
kindly  verified  the  identity  of  Cuthona  nana.  Carol  Ann  Kearns  helped  draw  some 
of  the  figures.  My  wife,  Mary-Jane,  was  a  constant  source  of  encouragement.  The 
support  of  a  University  of  New  Hampshire  Graduate  Fellowship  is  gratefully 
acknowledged. 

SUMMARY 

1.  The  larval  development,  metamorphosis,  and  postlarval  growth  of  the  eolid 
nudibranch,    Cuthona    nana,   is   described.     Hatching  occurred   within    19   days   at 
11-13°  C.    The  lecithotrophic  veligers  remained  nonpelagic  and  proceeded  to  meta- 
morphose within  another  two  days. 

2.  Adult  nutrition  did  not  affect  egg  size  or  subsequent  development  and  meta- 
morphosis. 

3.  Embryogenesis,  hatching,  and  metamorphosis  were  unaffected  by  the  presence 
or  absence  of  the  adult  nudibranch's  prey,  the  hydroid  Hydractinia  cchinata. 

4.  Different   temperatures   altered   the   rate   of   development   and   of   metamor- 
phosis but  not  the  type  of  development.     Egg  masses  collected  in  the  field  and  in- 
cubated at  the  temperature  at  which  they  were  collected  invariably  produced  non- 
pelagic  lecithotrophic  veligers  which  then  metamorphosed. 


174  BRIAN  R.  RIVEST 

5.  Newly  metamorphosed  specimens  of  C.  nana  survived  for  up  to  six  weeks  at 
11-13°  C  and  ten  weeks  at  4°  C  in  the  absence  of  //.  cchinata. 

6.  In  the  presence  of  abundant  food,  specimens  of  C.  nana  deposited  fertile  egg 
masses  within   11    weeks  after  metamorphosis  at   11-13°   C,  and  continued  feeding 
and  ovipositing  for  two  months. 

7.  CntJiona  nana  feeds  specifically  on  Hydractinia  cchinata,  which  in  (josport 
Harbor  is  found  predominantly  on  shells  occupied  by  Pai/nnts  acadianns.     As  the 
hermit  crabs  move  about,  postlarvae  of  C.  nana  are  swept  up  by  the  gastrozooids  of 
//.  ccliinata,  are  not  eaten  by  the  polyps  but  reorient  and  feed  on  hydroid  tissue. 

8.  Nonpelagic  development  in  C.  nana  appears  to  result  in  a  patchy  distribution 
of  postlarvae  on  the  bottom,  and  an  uneven,  nonrandom  distribution  of  young  nudi- 
branchs  on  the  H.  cchinata  colonies. 

9.  Ciitlwna  nana  does  not  kill  the  //.  cchinata  colonies  it  preys  upon,  but  only 
crops  some  of  the  polyps  before  leaving  the  colony  to  find  a  mate  or  deposit  eggs. 
Lost  polyps  are  subsequently  regenerated. 

LITERATURE  CITED 

BAYNE,  B.  L.,   1972.     Some  effects  of  stress  in  the  adult  on  the  larval  development  of  Mytilus 

edulis.    Nature,  237  :  459. 
BAYNE,  B.  L.,  P.  A.  GABBOTT,  AND  ].  WIDDOWS,  1975.     Some  effects  of  stress  in  the  adult  on 

the  eggs  and  larvae  of  Mytilus  cdulis  L.  J.  Mar.  Biol.  Assoc.   U.K..  55:  675-689. 
BERRILL,  N.  J.,  1931.     The  natural  history  of  Bulla  hvdatis  Linn.  ./.  Mm:  Biol.  Assoc.  U.K.,  17: 

567-571. 
BONAR,  D.  B.,  1976.     Molluscan  metamorphosis:    a  study   in  tissue  transformation,  .tin.  Zool., 

16:  573-591. 
BONAR,  D.  B.,  AND  M.  G.  HADFIELD,  1974.     Metamorphosis  of  the  marine  gastropod  Phcstilla 

sibogac  Bergh   (Nudibranchia :    Aeolidacea)   I.  Light  and  electron  microscopic  analysis 

of  larval  and  metamorphic  stages.     /.  E.r[>.  Mm:  Biol.  Ecol.,  16:  227-255. 

BUNTING,  M.,  1894.     The  origin  of  sex-cells  in  Hydractinia  and  f'odocorync ;  and  the  develop- 
ment of  Hydractinia.    J.  Morfihol.,  9:  203-236. 
CASTEEL,  D.  B.,  1904.     The  cell-lineage  and  early  larval  development  of  Fiona  marina,  a  nudi- 

branch  mollusc.    Proc.  A  cad.  Nat.  Sci.  Phila.,  56:  325-405. 
CAZAUX,   C.   1961.     Signification  et  origine  de  1'association  entre  Hydractinie  et   Pagure :   role 

des  tropismes  larvaires  dans  le  developpement  de  1'hydraire.     Bull.  Sta.  Biol.  Arcachon, 

13  :  1-5. 
CHRISTENSEN,  H.  E.,  1967.    Ecology  of  Hydractinia  cchina/a  (Fleming)    ( Hydroidea,  Athecata) 

I.  Feeding  biology.    Ophelia,  4  :  245-257. 

CHRISTENSEN,  H.  E.,  1977.     Feeding  and  reproduction  in  Prccittliona  pcachi   (Mollusca  Nudi- 
branchia).    Ophelia.  16:  131-142. 
FARRAN,   G.   P.,   1903.     The  nudibranchiate  molluscs   of   Ballynakill   and   Bonn   Harbours,    Co. 

Galway.   Refit.  Fish.  Ireland  for  1901,  pt.  2:  123-132.' 
FRANZ,   D.   R.,    1970.      Possible   variability   in    larval   development   between   populations   of   the 

cephalaspid  opisthobranch  Actcocina  canalicu/ata.    Am.  J\Ialacol.   Union   Inc.  Bull.,  37: 

68-69. 
FRANZ,  D.  R.,  1971.     Development  and  metamorphosis  of  the  gastropod  Actcocina  canaliculata 

(Say).     Trans.  Am.  Microsc.  Soc..  90:  174-182. 
GERLACH,  S.  A.,  D.  K.  EKSTR^M,  AXD  P.  B.  ECKARDT,  1976.     Filter  feeding  in  the  hermit  crab, 

Pai/urus  hcrnhardus.    Occologia.  24:   257-264. 
GRAHAM,  J.  J.,  1970.     Coastal  currents  of  the  western  Gulf  of  Maine.     Int.  Coniin.  Nortlnvcst 

Ail.  Fish.  Res.  Bull.,  7 :  19-31. 
GRANT,   W.    C.,   JR.,    1963.      Notes   of   the   ecology   and   behavior   of  the   hermit   crab,    Pac/urus 

acadianus.   Ecology,  44:  767-771. 

GRANT,  W.  C.,  JR.,  AND  P.  J.   PONTIER,   1973.     Fitness  in  the  hermit  crab  Pagurus  acadianus 
with  reference  to  Hydractinia  cchinata.    Bull.  Mt.  Desert  1st.  Biol.  Lab.,  13:  50-53. 


DF.VKI.OPMKXT   OF  CUTHOXA   XAXA  175 

GRAXT,  \V.  C.,  JR.,  AND  K.  ~\\.  UI.MKR,   1974.     Sliell   selection  and  aggressive  behavior  in  two 

sympatric  species  of  hermit  crabs.     Blol.  Bull.,  146:  32-43. 
HARRIS,  L.  G.,  L.  W.  WRIGHT,  AND  B.  R.  RIVEST,   1975.     Observations  on  the  occurrence  and 

biology  of  the  aeolid  nudibranch  Cuthona  nana  in  X*e\v  England  waters.     Veliger,  17  : 

264-268. 
HEXDLER,  G.,  AND  D.  R.  FRANZ,  1971.     Population  dynamics  and  life  history  of  Crcpiditla  con- 

I'c.ra  (Gastropoda:    Prosobranchia)   in  Delaware  Bay.    Biol.  Bull.,  141:  514-526. 
HYMAX,  L.  H.,  1940.     The  Inrerfehrates,  J'ol.  I'.    Protozoa  through  Ctcnophora.    McGraw-Hill 

Book  Company.  Inc..  XTew  York,  726  pp. 
JEXSEN,  K.,  1970.    The  interaction  between  Pagunts  bernhardus  (L.)  and  Hydractinia  echinata 

(Fleming).    Ophelia.  8  :  135-144. 
JENSEN,   K..   1975.     The   profit   to   Pagurus  bernhardiis    (L.)    by   the   presence   of  Hydractinia 

echlnata  (Fleming).    Biokon  Rep.,  1:  1-4. 
LODER,  T.,  F.  E.  ANDERSON,  AND  T.  C.   SHEVENELL,   1973.     Sea   monitoring  of  cmplaced  baled 

solid  waste.     University  of  New  Hampshire  Report,  U.XT.H.   SG-118,   107  pp. 
ORTON,  J.  H.,   1927.     On  the  mode  of  feeding  of  the  hermit  crab,  Eupagarus  bernhardus,  and 

some  other  Decapoda.     /.  Mar.  Biol.  Assoc.   U.K.,  14:  909-921. 
PELSENEER,  R.,  1911.     Reserches  sur  rembryologie  des  gastropodes.     Acad.  Roy.  Belg.  Cl.  Sci. 

Mem.  Collect.  (Quarto).  3:  1-167. 
RAO,    K.    V.,    1961.      Development    and    life    history    of    a    nudibranchiate    gastropod    Cuthona 

adyarcnsis  Rao.  /.  Mar.  Biol.  Assoc.  India,  3  :   186-197. 
RASMUSSEN,  E.,  1944.    Faunistic  and  biological  notes  on  marine  invertebrates.     I.  The  eggs  and 

larvae  of  Brachystomia  rissoides  (Hani.),  Eitlimella  nitidissima   (Mont.),  Ketusa  trun- 

catiila    (Brug.)    and  Emb/etonia  pallida    ( .\.  &  H.),    (Gastropoda  marina).     Vidcnsk. 

Maid.  Dan.  Naturliist.  Foren.,  107:  207-233. 
REES,   YV.  J.,   1967.     A  brief  survey  of  the  symbiotic  associations  of  Cnidaria  with   Mollusca. 

Proc.  Malacol  Soc.  London.  37:  213-231. 
ROGINSKAYA,  I.  S.,  1962.    Reproductive  biology  and  life  cycle  of  Cuthona  pustulata  (Gastropoda, 

Nudibranchia).     Dokl.  Akad.  Xauk.  SSSR  Set:'  Biol.,  146:  488-491. 

SCHITFSMA,   K.,    1935.     Observations   on   Hydractinia   echinata    (Flem.)    and   Eupagnrus  bern- 
hardus (L.).    Arch.  Xecrl.  Zool.,  l\  261-314. 
SPIC.HT,  T.  M.,  1975.     Factors  extending  gastropod  embryonic  development  and  their  selective 

cost.    Occologia,  21 :  1-16. 
SUTHERLAND,  J.  P.,  1975.     Life  histories  and  the  dynamics  of  fouling  communities.     Pages  137- 

153  in  J.  D.  Costlow,  Ed.,   The  ecology  of  jouling  communities.     U.S. — U.S.S.R.   Co- 
operative Program,  Duke  University  Marine  Laboratory,  Beaufort,  North  Carolina. 
SWENNEN,    C.,    1961.      Data   on   distribution,    reproduction   and   ecology   of   the   nudibranchiate 

molluscs  occurring  in  the  Netherlands.     Xcth.  J.  Sea  Res.,  1  :   191-240. 
TARDV,  J.,   1970.     Contribution  a  1'etude  des  metamorphoses  chez  les  nudibranches.    Ann.  Sci. 

Nat.  Zool.  Biol.  Anim.,  12:  299-370. 
THOMPSON,    T.    E.,    1958.      The   natural    history,    embryology,    larval    biology    and    post-larval 

development   of  Adlaria   pro.vima    (A.   &•    H.)     (Gastropoda,    Opisthobranchia).     Phil. 

Trans.  Roy.  Soc.  Land.  Scr.  B.  242:  1-58. 
THOMPSON,  T.  E.,   1962.     Studies  on  the  ontogeny  of   Tritonia  hombcrgi  Cuvier    (Gastropoda 

Opisthobranchia).     Phil.  Trans.  Roy.  Soc'.  Lond.  Ser.  B,  245:  171-218. 
THOMPSON,  T.  E.,  1967.     Direct  development  in  a  nudibranch,  Cadlina  lacris,  with  a  discussion 

of  developmental  processes  in  Opisthobranchia.   /.  Mm:  Biol.  Assoc.  U.K.,  47:  1-22. 
VAX  DE  VVVER,  G.,  1964.    Etude  histologique  du  developpement  $H \dractinia  echinata  (Flem.). 

Cah.  Biol.  Mar.,  5:  295-310. 
VAX    DE   VYVER,    G.,    1967.      Etude    du    developpement    embryonnaire    des    hydraires    athecates 

(gymnoblastiques)  a  gonophores.   Arch.  Biol.,  78:  451-518. 
WOBBKR,  D.  R.,  1970.    A  report  on  the  feeding  of  Dendronotus  iris  on  the  anthozoan  Ccrianthus 

sp.  from  Monterey  Bay,  California.    Vcligcr,  12:  383-387. 


Continued  from  Cover  Two 

4.  Literature  Cited.     The  list  of  references  should    be    headed    LITERATURE  CITED, 
should  conform  in  punctuation  and  arrangement  to  the  style  of  recent  issues  of  THE  BIOLOGICAL 
BULLETIN,  and  must  be  typed  double-spaced  on  separate  pages.     Note  that  citations  should 
include  complete  titles  and  inclusive  pagination.     Journal  abbreviations  should  normally  follow 
those  of  the  U.  S.  A.  Standards  Institute  (USASI),  as  adopted  by  BIOLOGICAL  ABSTRACTS  and 
CHEMICAL  ABSTRACTS,  with  the  minor  differences  set  out  below.     The  most  generally  useful  list 
of  biological  journal  titles  is  that  published  each  year  by  BIOLOGICAL  ABSTRACTS  (BIOSIS  List  of 
Serials;   most  recent  issue,    1976).     Foreign   authors,  and  others  who  are  accustomed   to   use 
THE  WORLD  LIST  OF  SCIENTIFIC  PERIODICALS,  may  find  a  booklet  published  by  the  Biological 
Council  of  the  U.K.  (obtainable  from  the  Institute  of  Biology,  41  Queen's  Gate,  London,  S.W.7, 
England,  U.K.  at  £0.65  or  $1.75)  useful,  since  it  sets  put  the  WORLD  LIST  abbreviations  for  most 
biological  journals  with  notes  of  the  USASI  abbreviations  where  these  differ.     CHEMICAL  AB- 
STRACTS publishes  quarterly  supplements  of  additional  abbreviations.     The  following  points  of 
reference  style  for  THE  BIOLOGICAL  BULLETIN  differ  from  USASI   (or  modified  WORLD  LIST) 
usage: 

A.  Journal  abbreviations,  and  book  titles,  all  underlined  (for  italics) 

B.  All  components  of  abbreviations  with  initial  capitals  (not  as  European  usage  in  WORLD 
LIST  e.g.  J.  Cell.  Comp.  Physiol.  NOT  /.  cell.  comp.  Physiol.) 

C.  All  abbreviated  components  must  be  followed  by  a  period,  whole  word  components 
must  not  (not  strictly  as  USASI  usage,  i.e.  J.  Cancer  Res.) 

D.  Space  between  all  components  (e.g.  J.  Cell.  Comp.  Physiol.  not  J.Cell.Comp. Physiol.) 

E.  We  strongly  recommend  that  more  unusual  words  in  journal  titles  be  spelled  out  in  full, 
rather  than  employing  lengthy,  peculiar  "abbreviations"  or  new  abbreviations  invented  by  the 
author.     For  example,  use  Kit  Visindafjelags  Islendinga  without  abbreviation.     Even  in  more 
familiar  languages,  Z.   Vererbungslehre  is  preferred  to  Z.   VererbLehre  (WORLD  LIST)  or  Z.  Verer- 
bungsl.   (USASI).     Accurate  and  complete  communication  of  the  reference  is  more  important  than 
minor  savings  in  printing  costs. 

F.  All  single  word  journal  titles  in  full  (e.g.  Veliger,  Ecology,  Brain). 

G.  The  order  of  abbreviated  components  should  be  the  same  as  the  word  order  of  the  com- 
plete title  (i.e.  Proc.  and  Trans,  placed  where  they  appear,  not  transposed  as  in  some  BIOLOGICAL 
ABSTRACTS  listings). 

H.  Spell  out  London,  Tokyo,  Paris,  Edinburgh,  Lisbon,  etc.  where  part  of  journal  title. 
I.  Series  letters  etc.  immediately  before  volume  number. 

J.  A  few  well-known  international  journals  in  their  preferred  forms  rather  than  WORLD 
LIST  or  USASI  usage  (e.g.  Nature,  Science,  Evolution  NOT  Nature,  Land.;  Science,  N.  Y.;  Evolution, 
Lancaster,  Pa.) 

K.  The  correct  abbreviation  for  THE  BIOLOGICAL  BULLETIN  is  Biol.  Bull. 

5.  Figures.     The  dimensions  of  the  printed  page,  5  by  7f  inches,  should  be  kept  in  mind  in 
preparing  figures  for  publication.     Illustrations  should  be  large  enough  so  that  all  details  will  be 
clear  after  appropriate  reduction.     Explanatory  matter  should  be  included  separately  in  legends 
as  far  as  possible,  although  the  axes  should  always  be  numbered  and  identified  on  the  illustration 
itself.     Figures  should  be  prepared  for  reproduction  as  either  line-cuts  or  halftones;  no  other 
methods  will  be  used.     Figures  to  be  reproduced  as  line-cuts  should  be  drawn  in  black  ink  on  white 
paper,  good  quality  tracing  cloth  or  plastic,  or  blue-lined  coordinate  paper;  those  to  be  reproduced 
as  halftones  should  be  mounted  on  board,  and  both  designating  numbers  or  letters  and  scale-bars 
should  be  affixed  directly  on  the  figures.     We  recommend  that  halftones  submitted  to  us  be 
mounted  prints  made  at  about  1?  times  the  linear  dimensions  of  the  final  printing  desired  (the 
actual  best  reductions  are  achieved  from  copy  in  the  range  from  1$  to  2  times  the  linear  dimen- 
sions).    As  regards  line-blocks,  originals  can  be  designed  for  even  greater  reductions  but  are  best 
in  the  range  \.\  to  3  times.     All  figures  should  be  numbered  in  consecutive  order,  with  no  distinc- 
tion between  text  and  plate-figures.     The  author's  name  should  appear  on  the  reverse  side  of  all 
figures,  and  the  inked  originals  for  line-blocks  must  be  submitted  for  block-making. 

6.  Mailing.     Manuscripts  should  be  packed  flat.     All  illustrations  larger  than  8 J  by  11  inches 
must  be  accompanied  by  photographic  reproductions  or  tracings  that  may  be  folded  to  page  size. 

Reprints.  Reprints  may  be  obtained  at  cost;  approximate  prices  will  be  furnished  by  the 
Managing  Editor  upon  request. 


CONTENTS 


ANDERSON,  JOHN  MAXWELL 

Studies  on  functional  morphology  in  the  digestive  system  of  Oreaster 
reticulatus  (L.)  (Asteroidea) 1 

ARMSTRONG,  DAVID  A.,  DEBBIE  CHIPPENDALE,  ALLEN  W.  KNIGHT  AND 

JOHN  E.  COLT 

Interaction  of  ionized  and  un-ionized  ammonia  on  short-term 
survival  and  growth  of  prawn  larvae,  Macrobrachium  rosenbergii.  .  15 

BARKER,  M.  F. 

Descriptions  of  the  larvae  of  Stichaster  australis  (Verrill)  and 
Coscinasterias  calamaria  (Gray)  (Echinodermata :  Asteroidea) 
from  New  Zealand,  obtained  from  laboratory  culture 32 

CONKLIN,  D.  E.  AND  L.  PROVASOLI 

Biphasic  particulate  media  for  the  culture  of  filter  feeders 47 

GOVIND,  C.  K.  AND  FRED  LANG 

Development  of  the  dimorphic  claw  closer  muscles  of  the  lobster 
Homarus  americanus.  III.  Transformation  to  dimorphic  muscles  in 
juveniles "•  •  55 

GREEN,  JEFFREY  D. 

The  annual  reproductive  cycle  of  an  apodous  holothurian,  Lepto- 
synapta  tenuis:  a  bimodal  breeding  season 68 

HENDLER,  GORDON 

Development  of  Amphioplus  abditus  (Verrill)  (Echinodermata: 
Ophiuroidea).  II.  Description  and  discussion  of  ophiuroid  skeletal 
ontogeny  and  homologies 79 

HOVE,  H.  A.  TEN  AND  J.  C.  A.  WEERDENBURG 

A  generic  revision  of  the  brackish-water  serpulid  Ficopomatus 
Southern  1921  (Polychaeta :  Serpulinae),  including  Mercierella 
Fauvel  1923,  Sphaeropomatus  Treadwell  1934,  Mercierellopsis 
Rioja  1945  and  Neopomatus  Pillai  1960 96 

KURIS,  ARMAND  M. 

Life  cycle,  distribution  and  abundance  of  Carcinonemertes  epialti,  a 
nemertean  egg  predator  of  the  shore  crab  Hemigrapsus  oregonensis, 
in  relation  to  host  size,  reproduction,  and  molt  cycle 121 

MICKEL,  T.  J.  AND  J.  J.  CHILDRESS 

The  effect  of  pH  on  oxygen  consumption  and  activity  in  the  bathy- 
pelagic  mysid  Gnathophausia  ingens 138 

PEZALLA,  PAUL  D.,  ROBERT  M.  DORES  AND  WILLIAM  S.  HERMAN 

Separation  and  partial  purification  of  central  nervous  system  peptides 
from  Limulus  polyphemus  with  hyperglycemic  and  chromatophoro- 
tropic  activity  in  crustaceans 148 

RIVEST,  BRIAN  R. 

Development  of  the  eolid  nudibranch  Cuthona  nana  (Alder  and 
Hancock,  1842),  and  its  relationship  with  a  hydroid  and  hermit 
crab.,  157 


\ 


Volume  154 


F^Y 

AY  9 


oods  Hole,  Mass. 


Sumber  2 


BIOLOGICAL  BULLETIN 


PUBLISHED  BY 
THE  MARINE  BIOLOGICAL  LABORATORY 

Editorial  Board 


EDWARD  M.  BERGER,  Dartmouth  College 
JOHN  M.  ANDERSON,  Cornell  University 
JOHN  B.  BUCK,  National  Institutes  of  Health 

JOHN  D.  COSTLOW,  Duke  University 
PHILIP  B.  DUNHAM,  Syracuse  University 
J.  B.  JENNINGS,  University  of  Leeds 


MEREDITH  L.  JONES,  Smithsonian  Institution 

HOWARD  A.  SCHNEIDERMAN,  University  of 

California,  Irvine 

RALPH  I.  SMITH,  University  of  California, 

Berkeley 

F.  JOHN  VERNBERG,  University  of 

South  Carolina 

CARROLL  M.  WILLIAMS,  Harvard  University 


W.  D.  RUSSELL-HUNTER,  Syracuse  University 
Managing  Editor 


APRIL,  1978 


Printed  and  Issued  by 
LANCASTER  PRESS,  Inc. 

PRINCE  &  LEMON  STS. 
LANCASTER,  PA. 


THE  BIOLOGICAL  BULLETIN 

THE  BIOLOGICAL  BULLETIN  is  published  six  times  a  year  by  the  Marine  Biological  Laboratory, 
MBL  Street,  Woods  Hole,  Massachusetts  02543. 

Subscriptions  and  similar  matter  should  be  addressed  to  THE  BIOLOGICAL  BULLETIN,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts.  Agent  for  Great  Britain:  Wheldon  and 
Wesley,  Limited,  2,  3  and  4  Arthur  Street,  New  Oxford  Street,  London,  W.  C.  2.  Single  numbers, 
$8.00.  Subscription  per  volume  (three  issues),  $22.00,  (this  is  $44.00  per  year  for  six  issues). 

Communications  relative  to  manuscripts  should  be  sent  to  Dr.  W.  D.  Russell-Hunter,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts  02543  between  June  1  and  September  1,  and 
to  Dr.  W.  D.  Russell-Hunter,  P.O.  Box  103,  University  Station,  Syracuse,  New  York  13210, 
during  the  remainder  of  the  year. 


Copyright   ©  1978,  by  the  Marine  Biological  Laboratory 
Second-class  postage  paid  at  Woods  Hole,  Mass.,  and  additional  mailing  offices. 


INSTRUCTIONS  TO  AUTHORS 

THE  BIOLOGICAL  BULLETIN  accepts  original  research  reports  of  intermediate  length  on  a  variety 
of  subjects  of  biological  interest.  In  general,  these  papers  are  either  of  particular  interest  to  workers 
at  the  Marine  Biological  Laboratory,  or  of  outstanding  general  significance  to  a  large  number  of 
biologists  throughout  the  world.  Normally,  review  papers  (except  those  written  at  the  specific 
invitation  of  the  Editorial  Board),  very  short  papers  (less  than  five  printed  pages),  preliminary 
notes,  and  papers  which  describe  only  a  new  technique  or  method  without  presenting  substantial 
quantities  of  data  resulting  from  the  use  of  the  new  method  cannot  be  accepted  for  publication.  A 
paper  will  usually  appear  within  four  months  of  the  date  of  its  acceptance. 

The  Editorial  Board  requests  that  manuscripts  conform  to  the  requirements  set  below; 
those  manuscripts  which  do  not  conform  will  be  returned  to  authors  for  correction  before  review 
by  the  board. 

1.  Manuscripts.     Manuscripts  must  be  typed  in  double  spacing  (including  figure  legends, 
foot-notes,  bibliography,  etc.)  on  one  side  of  16-  or  20-lb.  bond  paper,  8?  by  11  inches.     They 
should  be  carefully  proof-read  before  being  submitted  and  all  typographical  errors  corrected 
legibly  in  black  ink.     Pages  should  be  numbered.     A  left-hand  margin  of  at  least   1^  inches 
should  be  allowed. 

2.  Tables,  Foot-Noles,  Figure  Legends,  etc.     Tables  should  be  typed  on  separate  sheets  and 
placed  after  the  Literature  Cited.     Because  of  the  high  cost  of  setting  such  material  in  type 
authors  are  earnestly  requested  to  limit  tabular  material  as  much  as  possible.     Similarly,  foot- 
notes to  tables  should  be  avoided  wherever  possible.     If  they  are  essential,  they  should  be  indi- 
cated by  asterisks,  daggers,  etc.,  rather  than  by  numbers.     Foot-notes  are  not  normally  permitted 
in  the  body  of  the  text.     Such  material  should  be  incorporated  into  the  text  where  appropriate. 
Explanations  of  figures  should  be  typed  double-spaced  and  placed  on  separate  sheets  at  the  end 
of  the  paper. 

3.  A  condensed  title  or  running  head  of  no  more  than  35  letters  and  spaces  should  be  included. 

4.  Literature  Cited.     The  list  of  references  should  be  headed  LITERATURE  CITED, 
should  conform  in  punctuation  and  arrangement  to  the  style  of  recent  issues  of  THE  BIOLOGICAL 
BULLETIN,  and  must  be  typed  double-spaced  on  separate  pages.     Note  that  citations  should 
include  complete  titles  and  inclusive  pagination.     Journal  abbreviations  should  normally  follow 
those  of  the  U.  S.  A.  Standards  Institute  (USASI),  as  adopted  by  BIOLOGICAL  ABSTRACTS  and 
CHEMICAL  ABSTRACTS,  with  the  minor  differences  set  out  below.     The  most  generally  useful  list 

Continued  on  Cover  Three 


Vol.  154,  No.  2  April,  1978 

THE 

BIOLOGICAL  BULLETIN 

PUBLISHED   BY   THE    MARINE   BIOLOGICAL   LABORATORY 


INCREASE  IN  RANGE  OF  TEMPERATURE  TOLERANCE  BY 
ACCLIMATION  IN  THE  COPEPOD 

EURYTEMORA  AFFINIS 

Reference:  B'wl  Bull,  154:  177-187.     (April,  1978) 

BRIAN  P.  BRADLEY 

Department  of  Biological  Sciences,  University  of  Maryland  Baltimore  County, 

Catonsvillc,  Maryland  21228 ' 

Adaptation  to  temperature  change  is  an  obvious  requirement  for  the  survival  of 
a  temperate  species  whose  habitat  is  subject  to  seasonalities.  The  response  of  a 
population  to  environmental  stress  depends  on  the  time  and  intensity  of  the  stress. 
Slobodkin  and  Rapoport  (1974)  suggest  that  if  one  level  of  response  (for  example 
physiological)  is  sufficient  to  meet  the  environmental  challenge,  the  next  level  (for 
example  change  in  gene  frequency )  need  not  be  invoked. 

The  calanoid  copepod,  Eurytcniora  affinis  (Poppe),  found  in  the  Chesapeake 
Bay  in  temperatures  ranging  from  0  to  30°  C,  has  many  generations  per  year; 
hence,  it  could  adapt  to  this  variability  in  temperature  either  on  an  individual  or 
population  level.  If  the  range  of  individual  tolerance  were  sufficient,  the  species 
might  not  need  to  change  genetically  through  the  year.  Bradley  (1975)  found  that 
individual  copepods  could  tolerate  the  0-30°  C  range  for  short  periods. 

The  question  addressed  in  this  paper  is  whether  individuals  can  become  adapted 
physiologically  to  a  wider  range  of  temperatures  (acclimation).  This  paper  also 
explores  the  effects  of  temperature,  salinity,  sex  and  stage  of  development  on  ac- 
climation to  high  temperature,  the  effects  of  temperature  and  sex  on  acclimation  to 
low  temperature,  and  the  relationship  between  tolerances  to  high  and  low  tempera- 
ture. 

MATERIALS  AND  METHODS 

Specimens  of  Eurytemora  used  in  some  of  the  experiments  were  descended  from 
animals  collected  from  Bear  Creek  in  the  upper  Chesapeake  Bay  in  winter.  In 
other  experiments,  on  salinity  effects  and  on  survival,  the  animals  originated  in  the 
middle  reaches  of  the  Patuxent  River,  Maryland,  in  late  spring.  The  reason  for 

177 

Copyright  ©  1978,  by  the  Marine  Biological  Laboratory 
Library  of  Congress  Card  No.  A38-518 


178  BRIAN  P.  BRADLEY 

the  different  collection  sites  was  the  availability  of  specimens.  Only  one  source  of 
animals  was  used  for  each  experiment.  The  experiments  on  heat  tolerance,  except 
those  on  salinity  effects,  were  done  in  water  from  the  Patuxent  River  with  near 
zero  salinity.  Those  on  cold  tolerance  were  done  in  5%c  water  from  Bear  Creek. 
Salinity  was  measured  with  a  refractometer. 

Acclimation  is  defined  in  this  paper  as  the  increase  in  temperature  tolerance  of 
individuals  following  exposure  to  a  temperature  closer  to  the  extreme  temperature, 
whether  high  or  low. 

Tolerance  to  high  temperature  was  measured  using  the  shock-recovery  assay 
(Bradley,  1975),  permitting  data  to  lie  obtained  on  individual  copepods.  The 
measurements  were  highly  repeatahle  with  test-retest  correlations  of  around  0.8, 
and  were  also  closely  related  to  survival  time  at  high  temperatures  (Bradley,  1976). 
Single  animals  in  2  ml  of  water  in  test-tuhes  were  immersed  in  an  aquarium  held 
constant  at  34.5°  C  using  a  heating-stirring  unit.  No  temperature  gradients  in  the 
aquarium  were  detected,  and  the  temperature  in  the  test-tubes  reached  34.5°  C 
within  90  seconds.  Time  to  succumb  (TS),  or  enter  a  coma,  and  time  to  recover 
(TR)  were  observed  during  a  30  min  exposure  of  all  test  animals  to  34.5°  C. 
Animals  were  considered  comatose  when  rotating  and  agitating  the  vial  failed  to 
rouse  them.  Recovery  was  noted  at  the  first  movement.  The  assay  period  was  30 
min,  and  the  measures  of  tolerance  were  combined  in  an  index  30  +  TS--TR, 
which  could  range  from  0  to  60,  the  higher  number  indicating  the  greatest  toler- 
ance. All  animals  were  exposed  to  the  34.5°  C  temperature  for  30  min,  and  those 
failing  to  recover  while  exposed  to  this  temperature  were  given  a  TR  of  30--  TS, 
the  index  becoming  2TS  in  these  cases.  No  attempts  were  made  to  distinguish 
between  coma  and  lethality,  but  both  TS  alone  and  the  index  (30 -f  TS  -  TR) 
were  closely  related  to  survival  time  at  30°  C  and  higher  (Bradley,  1976). 

Similar  methods  were  used  to  test  tolerance  to  low  temperatures.  In  the  latter 
case,  animals  were  placed  in  an  aquarium  at  0.5°  C,  and  all  were  removed  after  10 
min  (whether  succumbed  or  not).  A  majority  of  animals  did  become  comatose 
before  10  min,  and  recovery  could  be  more  easily  observed  at  room  temperature. 
Animals  not  succumbing  at  all  were  arbitrarily  scored  50,  the  remainder  using  the 
index  described  above.  The  next  largest  score  was  40  (animals  succumbing  at  10 
min,  recovering  immediately),  since  animals  became  comatose  at  or  before  10  min 
or  not  at  all,  so  setting  the  maximum  tolerance  at  50  rather  than  60  reduced  the 
discontinuity  of  tolerance  scores. 

In  the  present  study  12  animals  and  all  treatment  groups  were  included  in  each 
run.  Each  set  of  experiments  was  done  by  the  same  observer.  In  the  case  of  heat 
tolerance,  variance  between  replicate  runs  was  treated  as  error  variance,  since  no 
interactions  between  treatment  and  run  were  detected.  The  net  effect  of  ignoring 
runs  was  to  make  the  tests  of  variables  more  conservative  because  of  the  increased 
error  variance.  In  the  case  of  cold  tolerance,  differences  between  runs  were  quite 
large,  due  to  difficulties  in  controlling  the  low  temperature  at  exactly  0.5°  C.  So 
run  variance  (and  interaction)  were  included  in  the  analyses  of  variances  of  cold 
tolerances. 

Tolerance  was  also  observed  as  longevity  in  two  constant  temperatures  and  in  a 


ACCLIMATION  IN  A  COPEPOD 


179 


slowly  increasing  temperature.  Acclimation  was  observed  in  these  ca>e.s  as  the  in- 
crease in  duration  of  activity  in  the  test  temperature  regime  of  animals  previously 
exposed  to  an  intermediate  temperature.  Half  the  animals  tested  in  increasing 
temperatures  were  maintained  at  24°  C  for  24  hr  before  the  temperature  \vas  raised 
from  24°  C  to  31°  C  in  30  min.  The  remainder  were  kept  at  15°  C.  Only  males 
were  included  in  these  experiments  in  increasing  temperatures.  All  the  animals 
were  then  exposed  to  a  temperature  of  31°  C  initially,  which  was  increased  1°  C 
every  30  min.  The  test  animals  were  continuously  monitored  and  the  times  when 
each  animal  succumbed  and  could  not  be  roused  were  noted.  In  this  case,  the  end- 

TABLE  I 

Increased  temperature  tolerance  of  animals  raised  at  10°  C  and  exposed  to  18°  C  and  24°  C  for  three 
periods  of  time.  Body  of  table  gives  mean  tolerances  (in  min)  to  high  temperature  measured  by  the 
shock-recovery  assay  described  in  Methods. 


Mean  temperature  tolerances 


Time  of  exposure 


2  day 

4  day 

7  day 

Exposure  temperature 
10°  C  (control)      9 

7.5 

9.3 

5.8 

c? 

5.8 

7.0 

5.8 

18°C 

9 

14.0 

12.8 

7.3 

0* 

9.0 

8.8 

10.8 

24°  C 

9 

29.0 

40.8 

36.6 

<? 

21.0 

22.8 

16.5 

8  animals  per  mean,  144  total 

Variance  analyses  for  each  sex 


Mean  squares 


Females 

Males 

Time  of  exposure 
Days  at     10°  C 
18°  C 

24.5 
63.2 

8.3 
19.0 

24°  C 

284.3 

166.4* 

Temperature  of  exposure 
18°  C,  24°  C  vs.  10°  C 
18°  C  vs.  24°  C 

3792.5* 

7575.2* 

1190.3* 
1344.1* 

Within  subclass 

159.7 

21.0 

Total  variances 

312.3 

57.1 

*  P  <  0.01. 


180  BRIAN  P.  BRADLEY 

TABU.  1 1 

Increased  high  lc»i/>c>-(it/irc  tolerances  of  males  raised  at  20°  C  compared  to  10°  C  and  exposed  to  three 
temperatures  for  two  days.  Body  of  table  gives  tolerances  measured  by  the  shock-recovery  assay.  All 
animals  recovered  when  raised  at  20°  C  and  two  of  24  recovered  when  raised  at  10°  C. 

Mean  temperature  tolerance 

Raised  at  10°  C  Raised  at  20°  C 

Exposure  temperature 

10°  C  6.7  31.7 

18°  C  10.0  43.4 

24°  C  25.0  47.3 

8  animals  per  mean         9  animals  per  mean 

Variance  analyses 

Mean  squares 


Raised  at  10°  C  Raised  at  20°  C 


Exposure  temperature 

18°  C,  24°  C  vs.  10°  C 
18°  C  vs.  24°  C 

622.0* 
900.0* 

1117.9* 
68.4 

Within  subclass 

39.4 

102.5 

Total  variances 

102.2 

140.7 

*  P  <  0.01. 

point  may  not  have  been  death  itself,  but  tantamount  to  death,  since  recovery  did 
not  occur  in  the  increasing  temperature. 

The  relationships  between  heat  and  cold  tolerance  were  measured  as  correlations 
between  observations  on  the  same  animals  assayed  for  cold  tolerance  and  heat 
tolerance  5-6  hr  apart  on  one  day.  Both  assays  were  repeated  the  next  day,  thus 
two  assays  for  heat  tolerance  and  two  for  cold  tolerance  were  done  on  each  animal. 

When  the  data  in  each  of  the  experiments  were  analyzed,  the  sexes  were  treated 
separately.  This  was  done  because  of  the  observed  differences  between  the  means 
and  variances  of  temperature  tolerances  of  the  two  sexes. 

RESULTS 

Acclimation  to  increased  temperature  occurs  in  Eurytemora  affinis  (Table  I). 
The  set  of  animals  exposed  to  18°  C  or  24°  C  prior  to  testing  were  significantly 
more  tolerant  than  those  kept  at  10°  C.  their  rearing  temperature.  The  largest 
effect  was  clearly  in  animals  exposed  to  24°  C,  since  they  were  significantly  more 
tolerant  than  those  exposed  to  18°  C.  Time  of  exposure  had  a  relatively  small 
effect,  although  it  was  significant  in  males  exposed  to  24°  C.  Females  appeared  to 
acclimate  more  than  males,  even  proportionally.  This  can  be  seen  from  the  changes 


ACCLIMATION  IN  A  COPEPOD 


181 


in  mean  tolerance,  especially  at  24°  C.  Furthermore,  of  the  16  animals  (out  of 
144)  recovering  within  30  min  of  the  temperature  shock  or  failing  to  succumb  at  all, 
14  were  female  and  2  were  male.  Females  also  seem  to  be  more  subject  to  en- 
vironmental influences  other  than  exposure  temperatures,  as  indicated  by  the  vari- 
ances within  treatments.  These  variances  were  159.7  for  females  and  21.0  for 
males.  The  greater  variance  between  females  is  consistent  with  their  greater  re- 
sponse to  exposure  temperature. 

By  comparison  with  the  low  rates  of  recovery  in  animals  raised  at  10°  C 
(above),  animals  raised  at  20°  C  almost  always  recovered  from  the  temperature 
shock  (Table  II).  In  this  experiment  on  rearing  temperature,  progeny  from  the 
same  stock  as  above  were  raised  at  20°  C  and  tested  after  exposure  to  10,  18,  and 
24°  C  as  before.  Only  males  were  tested  in  this  case.  The  results  in  Table  II 
clearly  show  the  increase  in  tolerance  of  the  animals  raised  at  20°  C,  regardless  of 


TABLE  III 

Increased  high  temperature  tolerance  of  animals  raised  at  10°  C  and  exposed  to  18°  C  and  23°  C  for 
3  hr  and  20  hr.     Body  of  table  gives  tolerances  measured  by  the  shock-recovery  assay. 

Mean  temperature  tolerance 

Exposure  time 


3  hr 

20  hr 

Exposure  temperature 
10°  C  (control)      9 

10.0 

12.0 

c? 

9.0 

13.0 

18°  C 

9 

11.5 

20.5 

<? 

9.5 

14.0 

23  °C 

9 

15.0 

51.0 

rf 

15.5 

17.0 

4  animals  per  mean                    4  animals  per  mean 

Variance  analyses  for  each  sex  and  exposure  time 


Mean  squares 


9  at  3  hr 

c?  at  3  hr 

9  at  20  hr 

c?  at  20  hr 

Exposure  temperatures 
23°  C,  18°  C  vs.  10°  C 
23°  C  vs.  18°  C 

28.2 
24.5 

32.7 
72.0* 

1504.2** 
1860.5** 

10.7 
18.0 

Within  subclass 

7.9 

11.8 

83.5 

63.9 

Total  variances 

11.2 

19.2 

374.2 

54.9 

*  P  <  0.05. 
**  P  <  0.01. 


182 


BRIAN  P.  BRADLEY 


TABLE  IV 

Increased  high  temperature  tolerance  at  higher  salinities  following  acclimation  at  two  temperatures  and 
two  salinities  for  24  hr.  Body  of  table  gives  tolerances  measured  by  shock-recovery  assay.  Animals 
in  Experiment  C  were  shocked  at  33.5°  C;  the  others  at  34.5°  C. 


Mean  temperature  tolerances 


Experiment 


A 

B 

C 

Exposure 
Temp. 

Salinity 

Exposure 
Temp. 

Salinity 

Exposure 
Temp. 

Salinity 

0%c 

13  %o 

0%o 

13& 

0%<, 

13&, 

9 

20°  C 

14.8 

30.3 

13°  C 

12.3 

27.8 

11°  C 

14.4 

18.4 

cf 

11.0 

23.8 

8.3 

12.3 

5.8 

9.4 

9 

24°  C 

19.8 

27.8 

23°  C 

13.2 

24.2 

23°  C 

22.6 

33.0 

rf1 

13.8 

18.8 

8.0 

11.5 

11.6 

26.2 

4  animals  per  mean 

6  animals  per  mean 

5  animals  per  mean 

Variance  analyses  for  each  sex 


Mean  squares 


Experiment 

A 

B 

C 

9 

cT 

9 

d" 

9 

d" 

Between  temperatures 
Between  salinities 

6.3 
552.3** 

5.0 

315.1** 

12.0 

1053.4** 

2.0 
84.4 

649.8* 

259.2 

638.4* 
414.1* 

Interaction 

56.3 

60.1 

30.4 

0.4 

51.2 

151.3 

Within  subclass 

58.1 

28.0 

76.0 

32.0 

90.0 

86.9 

Total  variance 

87.5 

47.8 

113.7 

31.6 

126.3 

136.5 

*  P  <  0.05. 
**P  <  0.01. 


what  temperature  they  were  exposed  to  later.  Thus,  acclimation,  and  adaptation  to 
an  elevated  temperature,  can  occur  during  development,  and  the  effect  of  a  low  tem- 
perature during  development  cannot  be  completely  overcome  by  subsequent  ex- 
posure to  a  higher  temperature. 

Having  shown  that  acclimation  occurred,  even  during  development,  the  next 
question  was  whether  a  short  exposure  time  would  suffice.  Table  III  shows  that 
as  little  as  3  hr  at  23°  C  and  certainly  less  than  24  hr  were  required  for  acclimation 
to  occur.  There  is  some  evidence  that  males  acclimate  earlier  and  less  than  do 


ACCLIA1ATION  IN  A  COPEPOD  183 

females.    This  was  indicated  also  in  the  earlier  data  in  Table  I.    The  same  animals 
were  tested  each  time,  and  the  correlation  between  measured  tolerances  was  0  48 


All  the  experiments  reported  so  far  were  done  in  water  with  no  detectable 
salinity.  Temperature  tolerance  was  shown  earlier  to  increase  when  animals  col- 
lected at  Qf/co  were  placed  in  higher  salinity  (Bradley,  1975),  so  it  seemed  reason- 
able to  test  for  an  effect  of  salinity  on  acclimation.  The  results  of  three  experiments 
are  shown  in  Table  IV.  None  of  the  three  experiments  gave  much  indication  that 
acclimation  was  influenced  by  salinity.  In  the  first  two  experiments  (A  and  B) 
the  shock  temperature  was  too  high;  but  even  when  there  was  sufficient  variation 
in  tolerance  (C),  no  evidence  of  interaction  between  exposure  temperature  and 
salinity  was  found. 

The  data  in  Table  IV  cannot  be  directly  compared  to  those  in  Tables  I  and  II, 
since  the  source  of  the  animals  differed  and  the  experiments  in  Table  IV  were  done 
almost  9  months  later.  However,  tolerances  of  females  were  again  higher,  as  were 
the  variances  among  females  within  treatment  as  discussed  earlier. 

Additional  experiments  on  acclimation  to  high  temperatures  were  done  using 
two  other  criteria,  survival  times  at  constant  high  temperatures  and  times  until 
complete  inactivity  of  animals  in  slowly  increasing  temperatures.  In  the  former 
experiments,  females  survived  longer  than  males  at  both  32°  C  and  33°  C,  but  there 
was  no  evidence  of  acclimation  in  animals  exposed  to  25°  C  for  24  hr.  Mean  sur- 
vival times  of  females  ranged  from  9.3  to  13.0  hr  and  of  males  ranged  from  3.7  to 
9.5  hr,  depending  mainly  on  test  temperature. 

There  was  evidence  of  acclimation,  when  time  to  inactivity  was  the  criterion. 
Animals  exposed  to  24°  C  for  24  hr  remained  active  significantly  longer  (146-151 
min)  than  did  animals  kept  at  15  hr  (74—103  min),  when  tested  in  a  temperature 
increasing  slowly  from  31°  C.  The  temperatures  at  immobilization  were  35.3  to 
35.7°  C  and  32.9°  C,  respectively. 

The  reasons  for  the  inconsistency  between  these  two  experiments  is  not  clear. 
In  the  second  set  of  experiments  the  total  time  of  observation  was  less  than  2.5  hr, 
allowing  a  more  accurate  measurement  of  longevity.  The  stress  due  to  tempera- 
ture probably  was  greater  in  the  second  experiments,  perhaps  allowing  more  ac- 
curate expression  of  the  effects  of  acclimation. 

Acclimation  to  decreased  temperature  also  occurs  in  Eurytemora  affinis,  al- 
though less  rapidly  than  to  increased  temperatures.  Two  sets  of  data  were  obtained, 
one  set  from  animals  tested  for  tolerance  following  exposure  to  4,  10,  and  15°  C 
for  24  hr  and  a  second  set  from  different  animals  exposed  to  4,  10,  and  15°  C  for 
60  hr  (Table  V).  There  is  clear  evidence  for  acclimation  to  low  temperatures, 
especially  after  60  hr.  There  also  seems  to  be  more  acclimation  in  males  by  24  hr 
and  more  in  females  by  60  hr,  which  is  consistent  with  the  inference  from  Tables  I 
and  II  that  males  acclimate  earlier  and  less  than  do  females.  However,  the  sexual 
dimorphism  in  degree  of  acclimation  was  much  less  for  cold  tolerance  than  for  heat 
tolerance.  Finally,  the  four  variances  (mean  squares)  within  run  and  exposure 
temperature  in  Table  V  taken  as  crude  measures  of.  physiological  variance,  are 
consistent  with  the  more  immediate  and  smaller  flexibilitv  of  males. 


184 


BRIAN   P.   BRADLEY 


TABLE  V 

Increased  low  temperature  tolennice  of  animals  raised  at  15°  C  and  exposed  to  10°  C  and  4°  C;  one  set 
exposed  for  24  hr  and  another  for  60  hr.     Indices  of  tolerance  are  in  body  of  table. 

Mean  temperature  tolerance 

Exposure  time 


24  hr 

60  hr 

Exposure  temperature 
15°  C  (control)       9 

23.6 
20.9 

31.8 
22.8 

10°  C                       9 

24.6 

27.3 

35.4 
23.8 

4°C                       9 

28.3 
29.2 

50.0 
35.3 

16  per  mean 

16  per  mean 

Variance  analyses  for  each  sex 


Mean  squares 


24  hr 

60  hr  exposure 

9 

d* 

9 

C? 

Exposure  temperatures 
4°  C  2)5.  10°  C,  18°  C 
10°  C  vs.  15°  C 
Runs 

187.1* 
8.0 

2084.8** 

276.8* 
331.6* 
522.4** 

3381.4** 
207.1 
137.1 

1928.0** 
9.1 

775.1** 

Runs  X  temps. 
Within  subclass 

Total  variance 

349.9** 
29.5 

121.5* 
51.3 

231.4 
138.6 

125.2 
75.8 

204.4 

101.1 

207.2 

160.2 

*  P  <  0.05. 
**  P  <  0.01. 


One  question  raised  by  these  results  is  whether  hot  and  cold  tolerances  are 
similar  or  different  characters.  Earlier  indications  (Bradley,  1975)  were  that 
animals  resistant  to  high  temperatures  tended  to  he  resistant  to  cold  temperatures. 
To  test  the  relationship  more  formally,  tests  of  heat  and  cold  tolerances  were  done 
twice  on  24  animals  on  successive  days.  All  the  correlations  except  one  were  posi- 
tive, eight  of  twelve  were  significant.  Test-retest  correlations  (heat-heat,  cold- 
cold)  averaged  0.45  for  males  and  0.70  for  females,  which  were  only  slightly  higher 
than  the  average  correlations  between  cold  and  heat  tolerances  (0.32  and  0.54, 
respectively).  Hence,  there  is  no  evidence  that  tolerance  was  biased  in  one  direc- 
tion in  each  animal.  In  other  experiments  where  both  tolerances  were  measured, 


ACCLIMATION  IN  A  COPEPOD  185 

correlations  were  sometimes  low  but  were  never  negative  when  averaged  over  the 
experiment. 


DISCUSSION 

Acclimation  to  high  temperatures  occurs  probably  quite  quickly  (<24  hr), 
being  completed  in  a  few  days,  perhaps  more  slowly  and  certainly  more  extensively 
in  females  than  in  males.  Exposure  to  increased  temperature  during  development 
also  leads  to  increased  tolerance  in  adults,  beyond  what  could  be  achieved  by  ex- 
posure beginning  at  the  adult  stage.  Apparently,  changes  affecting  temperature 
tolerance  occur  during  development  and  are  only  partially  reversible  in  the  adult 
stage.  Although  salinity  affects  temperature  tolerance,  there  is  no  evidence  that 
acclimation  is  greater  at  higher  salinity.  Acclimation  to  low  temperatures  also  oc- 
curs, but  less  rapidly  than  to  high  temperatures.  Effects  of  sex  on  acclimation  to 
cold  temperatures  are  also  less  marked. 

The  results  on  acclimation  to  high  temperatures  agree  with  the  observations  of 
others.  Levins  (1969)  found  that  most  of  the  thermal  acclimation  of  Drosophila 
species  took  place  in  the  first  12  hr.  Bowler  (1963a)  also  found  that  acclimation  in 
the  crayfish,  Astacus  pallipcs,  occurred  rapidly  and  was  completed  in  about  two  days. 
Yernberg  and  Moreira  (1974)  reported  that  males  of  the  copepod  species  Euter- 
pina  acutifrons  had  a  lower  metabolic  rate  at  15°  C  than  females  when  both  had 
been  acclimated  at  25°  C.  However,  males  were  smaller  and  the  metabolic  (respira- 
tion) rates  were  not  adjusted  for  body  size.  According  to  the  data  of  McLeese 
(1956),  from  a  study  of  the  effects  of  salinity,  acclimation,  and  oxygen  tension  on 
lobster  survival,  there  appeared  to  be  little  effect  of  salinity  on  acclimation. 

Data  in  this  study  also  suggest  that  acclimation  (at  least  to  high  temperature) 
is  more  easily  detected  using  coma  tolerance  rather  than  survival  as  the  criterion. 
Survival  time  was  not  increased  following  exposure  to  25°  C,  compared  with  15°  C. 
Where  coma  tolerance  was  the  criterion,  whether  in  a  shock  temperature  (Tables 
I— IV)  or  in  slowly  increasing  temperature,  the  data  indicate  that  significant  acclima- 
tion did  take  place.  Heinle  (1969)  also  reported  that  thermal  tolerances  of  E. 
affinis,  measured  as  survival  in  constant  environments,  was  not  increased  in  ani- 
mals exposed  to  20  or  25°  C,  compared  with  animals  exposed  to  10  or  15°  C. 
Hamby  (1975)  found  that  acclimation  of  a  marine  snail,  Littorina  lilt  or  ea,  sig- 
nificantly shifted  the  temperature  at  which  the  animal  entered  heat  coma  but  af- 
fected the  lethal  temperature  very  little.  He  concluded  that  the  nervous  system  of 
Litforina  was  most  vulnerable  to  thermal  extremes,  as  is  the  case  with  other 
poikilotherms  (Prosser,  1973). 

The  influence  of  acclimation  on  the  nervous  system  (and  so  on  coma  tolerance) 
is  indicated  by  the  results  of  Baldwin  and  Hochachka  (1970)  who  showed  that  dif- 
ferent variants  of  acetylcholinesterase  were  present  in  the  brains  of  trout  acclimated 
to  different  temperatures.  Other  reported  responses  to  exposure  to  higher  tem- 
perature were  lowered  temperature-specific  respiration  rates  in  a  toad  (Fitzpatrick 
and  Atebara,  1974),  lower  temperature-specific  respiration  rates  and  heart  rates 
in  limpets  (Markel,  1974),  and  alterations  in  enzyme  systems  donating  energy  re- 
quired in  the  functions  of  tissue  "cation  pumps"  (Bowler,  1963b). 


186  BRIAN  P.  BRADLEY 

Having  noted  the  agreement  with  other  results  and  descrihed  possible  mecha- 
nisms, the  question  remaining  is  how  such  ability  to  acclimate  (individual  flexi- 
bility) is  maintained  (or  how  it  arose),  when  no  individual  copepods  are  exposed 
to  the  whole  range  of  temperatures  in  the  Chesapeake  Bay  (0  to  30°  C).  Daily 
fluctuations  in  temperature,  together  with  diurnal  migration  may  be  sufficient  for 
physiological  flexibility  to  be  an  important  trait,  which  is  maintained  by  natural 
selection.  Another  (complementary)  hypothesis  is  that  tolerances  to  high  and  low 
temperatures  are  much  the  same  trait  genetically.  They  appear  to  be  related  pheno- 
typically,  as  shown  previously  (Bradley,  1975,  1976)  and  reported  again  in  this 
paper.  Thus,  the  flexibility  observed  may  be  the  result  of  natural  selection  for 
tolerance  to  extremes.  Such  selection  would  be  relaxed  in  intermediate  tempera- 
tures, but  never  reversed.  One  problem  with  this  explanation  is  that  in  several 
experiments  large  additive  genetic  components  of  variance  in  temperature  tolerance 
have  been  observed,  which  should  not  be  the  case  if  selection  is  always  in  the  same 
direction  (Bradley,  1978). 

Even  if  there  were  a  single  explanation  for  the  flexibility  observed,  the  reasons 
for  the  greater  flexibilities  or  acclimation  in  females  are  not  at  all  obvious.  Female 
specimens  of  Eurytciiiora  do  not  store  sperm  much  beyond  the  first  egg  sac  (Heinle 
and  Flemer,  1975),  although  such  storage  may  occur  occasionally.  If  males  are 
required  for  each  mating,  and  there  is  only  a  short  interval  between  fertilization 
and  hatching,  there  is  no  obvious  reason  why  males  should  be  less  tolerant  and  less 
flexible  than  females  at  high  temperatures. 


I  am  grateful  to  Dr.  Frank  Hanson  for  constructive  criticism  and  to  Richard 
Muths,  Jody  Myers,  Kenneth  Keeling,  Richard  Imbach,  Denise  Markoff,  and 
Margaret  Phelan  for  their  assistance.  Dr.  Ian  McLaren  of  Dalhousie  University 
commented  on  an  earlier  version  of  the  manuscript.  The  work  was  supported 
mainly  by  Grant  A-027  from  the  Annual  Allotment  Program,  Office  of  Water  Re- 
search and  Technology,  U.S.  Department  of  the  Interior.  Some  support  also  was 
derived  from  a  Matching  Grant  Agreement  with  OWRT  and  from  Grant  BMS-75- 
20282  from  the  National  Science  Foundation. 


SUMMARY 

The  copepod,  Eurytemora  affinis,  was  tested  for  its  ability  to  recover  from  short 
exposures  to  a  high  temperature  (temperature  tolerance).  Animals  kept  at  a  warm 
temperature  for  several  hours  or  days  before  the  test  increasd  in  tolerance  (ac- 
climation). Females  showed  higher  tolerance  and  acclimation  than  males.  Tem- 
perature tolerance  was  greater  at  a  higher  salinity  (13/{o  vs.  0%c),  but  acclimation 
was  not.  Analogous  tests  were  done  at  low  temperatures.  Acclimation  to  cold 
temperature  also  occurred,  but  more  slowly.  Sexual  differences  were  less  marked 
than  for  heat  tolerance.  When  tested  on  the  same  animals,  heat  and  cold  tolerances 
seemed  to  be  positively  related  traits. 


ACCLIMATION  IN  A  COPEPOD  187 

LITERATURE  CITED 

BALDWIN,  J.,  AND  P.  W.  HOCHACHKA,  1970.     Functional  significance  of  isoenzymes  in  thermal 

acclimation.     Acetylcholinestera.se  from  trout  brain.     Biochcm.  J.,  116:  883-887. 
BOWLER,  K.,   1963a.     A  study  of  factors  involved  in  acclimatization  to  temperature  and  death 

at  high  temperatures  in  Astaciis  pallipcs  1.     Experiments  on  intact  animals.     /.   Cell 

Comp.  Physiol.,  62  :  119-132. 
BOWLER,  K.,   1963b.     A  study  of  factors  involved  in  acclimatization  to  temperature  and  death 

at  high  temperatures  in  Astacus  pallipcs  2.     Experiments  at  the  tissue   level.  /.   Cell 

Comp.  Physiol,  62  :  133-146. 

BRADLEY,  B.  P.,  1975.     The  anomalous  influence  of  salinity  on  temperature  tolerances  of  sum- 
mer and  winter  populations  of  the  copepod  Eurytemora  affinis.  Biol.  Bull.,  148:  26-34. 
BRADLEY,  B.   P.,   1976.     The  measurement  of  temperature  tolerance :   verification  of  an  index. 

Limnol.  Oceanogr.,  21 :  596-599. 
BRADLEY,  B.  P.,  1978.     Genetic  and  physiological  adaptation  of  the  copepod  Ettrythemora  nffinis 

to  seasonal  temperatures.     Genetics,  in  press. 
FITZPATRICK,  L.  C.,  AND  M.  Y.  ATEBARA.  1974.     Effects  of  acclimation  to  seasonal  temperatures 

on  energy  metabolism  in  the  toad  Bufo  woodhousei.    Physiol.  ZooL,  47:  119-129. 
HAMBY,  R.  J.,  1975.     Heat  effects  on  a  marine  snail.    Biol.  Bull.,  149  :  331-347. 
HEINLE,  D.  R.,  1969.    Temperature  and  zooplankton.    Chesapeake  Set.,  10 :  186-209. 
HEINLE,  D.  R.,  AND  D.  A.  FLEMER,  1975.     Carbon  requirements  of  a  population  of  the  estuarine 

copepod,  Eurytemora  affinis.    Mar.  Biol.,  31  :  235-247. 
LEVINS,  R.,  1969.     Thermal  acclimation  and  heat  resistance  in  Drosophila  species.     Am.  Nat., 

103  :  483-499. 
MARKEL,   R.   P.,    1974.     Aspects   of  the   physiology   of   temperature   acclimation    in    the   limpet 

Acmaea  Umatula  Carpenter   (1964)  :  an  integrated  field  and  laboratory  study.  Physio!. 

ZooL,  47:  99-109. 
McLEESE,  D.  W.,  1956.     Effects  of  temperature,  salinity,  and  oxygen  on  the  survival  of  the 

American  lobster.   /.  Fish.  Res.  Board  Can.,  13 :  247-272. 

PROSSER,  C.  L.,  1973.     Comparative  animal  physiology,  3rd.  ed.    W.  B.  Saunders,  Co.,  Phila- 
delphia, 966  pp. 
SLOBODKIN,  L.  B.,  AND  A.  RAPOPORT,  1974.     An  optimal  strategy  of  evolution.     Q.  Rev.  Biol., 

49:   181-199. 
VERNBERG,  W.  B.,  AND  G.  S.  MOREIRA,  1974.    Metabolic-temperature  responses  from  the  copepod 

Euterpina  acutifrons  (Dana)  from  Brazil.    Comp.  Biochcm.  Physiol.,  49A :  757-761. 


Reference:  Biol  Bull,  154:  188-197.     (April,  1978) 


THE  INFLUENCE  OF  CONSTANT  AND  CYCLIC  ACCLIMATION  TEM- 
PERATURES ON  THE  METABOLIC  RATES  OF  PANOPEUS 
HERBSTII  AND  UCA  PU  GIL  AT  OR  l> 2 

R.  F.  DAME  3  AND  F.  J.  VERNBERG 

Belle  W.  Baruch  Institute  for  Marine  Biology  and  Coastal  Research,  University  of  South 
Carolina,  Columbia,  South  Carolina  2920S  USA 

Temperature  is  one  of  the  major  physical  factors  influencing  the  metabolic  rates 
of  intertidal  invertebrates  (Newell,  1975;  Vernberg  and  Vernberg,  1972).  Most 
previous  studies  on  the  respiratory  metabolism  of  intertidal  organisms  have  been 
conducted  at  constant  temperatures  and  have  utilized  organisms  acclimated  to  con- 
stant temperatures.  Although  these  studies  have  led  to  many  insights  into  the  in- 
fluence of  temperature  on  respiratory  adaptations,  they  may  not  describe  the  meta- 
bolic response  of  animals  subjected  to  fluctuating  thermal  environments  typical  of 
those  normally  encountered  in  nature.  Hence,  a  question  can  be  raised  concerning 
the  value  of  the  previous  data  from  these  studies  at  constant  temperature  in  the 
analysis  of  the  ecological  energetics.  To  understand  the  significance  of  respiration 
in  energy  transfer  in  an  ecosystem  requires  accurate  estimates  of  oxygen  consump- 
tion rates.  This  paper  reports  the  results  of  a  study  on  the  comparative  influence 
of  constant  and  cyclic  acclimation  temperatures  on  the  respiratory  metabolism  of 
two  intertidal  crabs  which  are  common  in  South  Carolina  estuaries,  the  mud  crab, 
Panopeus  herbstii  (Milne-Edwards),  and  the  fiddler  crab,  Uca  pugilator  (Bosc). 

Published  data  dealing  with  the  influence  of  cyclic  thermal  environments  on 
the  physiology  of  marine  invertebrate  animals  is  limited,  especially  for  respiratory 
metabolism.  Earlier  Kahn  (1965)  studied  the  effects  of  cyclic  temperature  on  the 
growth  of  copepods  while  observations  on  larval  crab  growth  under  cyclic  thermal 
regimes  were  reported  by  Costlow  and  Bookhout  (1971),  Christiansen  and  Costlow 
(1975),  and  Sastry  and  Vargo  (1977).  The  influence  of  cyclic  temperature  on 
survival  of  crab  larvae  (Costlow  and  Bookhout,  1971 ;  Sastry  and  Vargo,  1977)  and 
of  grass  shrimp  (Thorp  and  Hoss,  1975)  has  been  reported.  Sastry  and  Vargo 
(1977)  recorded  the  metabolic  response  of  larval  crabs  to  cyclic  temperatures,  as 
did  Humphreys  (1975)  for  the  nonmarine  wolf  spider.  Widdows  (1976)  and 
Bayne,  Widdows,  and  Worrall  (1977)  have  reported  on  the  influence  of  cyclic 
temperatures  on  the  physiology  of  a  bivalve,  Mytilus  edulis.  Some  investigators 
have  reported  that  the  metabolic  response  of  marine  animals  is  different  depending 
on  whether  the  temperature  was  increasing  or  decreasing  (Van  Winkle,  1969,  for 
the  mud  snail,  Nassarius  obsolctct;  Vernberg  and  Vernberg,  1966,  for  the  fiddler 
crab,  Uca  puyna.v}. 

1  This  research  was  supported  by  NSF  Grant  GA-3691S. 

2  Contribution  No.  194,  Belle  W.  Baruch  Institute  for  Marine  Biology  and  Coastal  Research, 
University  of  South  Carolina. 

3  Present  address :    Coastal  Carolina  College  of  the  University  of  South  Carolina,  Conway, 
South  Carolina  29526  USA. 

188 


CYCLIC  RESPIRATION  IN  CRABS 


189 


MATERIALS  AND  METHODS 

Mud  crabs,  Panopeus  licrhs/ii,  were  collected  from  intertidal  oyster  beds  in  the 
tidal  creeks  of  the  North  Inlet  Estuary  near  Georgetown,  South  Carolina.  Fiddler 
crabs,  Uca  piigilator,  were  collected  from  nearby  salt  marshes  where  they  are 
abundant  on  sand  beaches.  These  species  were  selected  because  each  has  distinctive 
ecological  requirements  and  both  are  present  in  great  numbers.  Thus,  it  is  pos- 
sible to  compare  responses  of  intertidal  crabs  from  different  habitats  to  determine 
any  commonality  of  response  by  intertidal  temperate  zone  animals. 

The  crabs  were  brought  into  the  laboratory,  washed  in  salt  water,  and  placed 
in  numbered  partitioned  plastic  boxes  containing  35%c  sea  water.  Then  the  boxes 
of  crabs  were  kept  in  Revco  Environmental  Chambers  in  which  light  and  tem- 
perature could  be  controlled.  The  crabs  were  fed  every  two  days,  and,  after  feed- 
ing, the  crabs  were  placed  in  a  clean  box  with  fresh  sea  water.  After  seven  days 
of  acclimation  to  a  constant  temperature,  the  respiration  rate  of  the  crabs  was 
determined  using  a  Gilson  Differential  Respirometer.  These  results  served  as  the 
baseline  value  against  which  measurements  determined  under  fluctuating  conditions 
were  compared.  Oxygen  consumption  was  computed  as  /A  Oi>/(hr-g  dry  weight), 
corrected  to  standard  temperature  and  pressure. 

After  completing  the  initial  metabolic  determinations,  the  environmental  cham- 
bers were  programmed  so  that  the  animals  would  experience  a  once  daily  cyclic 
temperature  regime  where  the  previous  constant  acclimation  temperature  was  the 
maximum  temperature  and  the  minimum  temperature  was  10°  C.  A  daily  10°  C 
thermal  change  was  selected  as  this  degree  of  fluctuation  is  not  uncommonly 
experienced  by  these  animals  throughout  much  of  the  year.  The  change  in  tem- 
perature followed  a  square  wave  with  about  an  hour  of  elapsed  time  before  a  new 
stabilized  temperature  was  reached.  After  thermal  cycling  had  started,  respiration 
rates  \vere  measured  on  days  3,  6,  9,  and  15-22.  At  the  end  of  the  experiments  the 
crabs  were  dried  in  an  oven  at  105°  C.  Although  different  photoperiods  have  been 
shown  to  influence  the  metabolic  response  of  crabs  (Dehnel,  1958),  the  photo- 
period  regime  was  the  same  for  animals  exposed  to  both  constant  and  fluctuating 
thermal  experiments. 

For  simplicity  of  experimental  design  and  to  compare  the  relative  effect  of  only 
thermal  regimes,  photoperiods  were  selected  which  corresponded  to  those  the 

TABLE  I 
The  varioiis  experimental  conditions  and  number  of  organisms  for  each  experiment. 


Cyclic  temperature 
range  (°  C) 

Photoperiod  (L:D) 

Constant  acclimation 
temperature 

Cyclic  acclimation 
(days) 

N 

5-15  (Panopeus) 

8:16 

15 

3,  6,9,  16 

27 

10-20  (Panopeus) 

12:12 

20 

3,  6,  9,  19 

25 

15-25  (Panopeus) 

14:10 

25 

3,  6,  9,  16 

25 

20-30  (Panopeus) 

14:10 

30 

3,  6,  9,  15 

27 

5-15  (Uca) 

8:16 

15 

3,  6,  9,  16 

28 

10-20  (Uca) 

12:12 

20 

3,  6,  9,  22 

24 

15-25  (Uca) 

14:10 

25 

3,  6,  9,  18 

25 

20-30  (Uca) 

14:10 

30 

3,  6,  9,  15 

23 

190 


R.  F.  DAME  AND  F.  J.  VERNBERG 


organisms  typically  experience  at  the  different  thermal  ranges.  For  example,  lo\v 
temperatures  normally  occur  during  periods  when  the  day  length  is  short ;  while,  in 
contrast,  long  days  and  high  temperatures  are  usually  coincident.  A  summary  of 
the  experimental  conditions  is  given  in  Table  I.  The  standard  statistical  techniques 
of  Steel  and  Torrie  (1960)  were  used  to  determine  means,  standard  errors,  and 
confidence  intervals. 

RESULTS 

Panopcus  herbstii  showed  no  statistically  different  change  in  its  respiratory 
response  at  the  low  and  high  temperature  ranges  of  5-15°  C  and  20-30°  C,  with 
oxygen  consumption  rates  the  same  after  constant  and  cyclic  acclimation  (Fig.  1 
and  Table  II).  However,  the  oxygen  consumption  response  to  the  middle  ac- 
climation temperatures  (10-20°  C  and  15-25°  C)  varied  with  acclimation  time. 
After  19  days  exposure  to  cyclic  temperature  acclimation,  the  10-20°  C  group  was 
consuming  about  the  same  amount  of  oxygen  as  it  did  after  acclimation  to  con- 
stant temperature.  However,  a  statistically  significant  (P  <  0.001)  decrease  in 
metabolic  rate  occurred  between  day  6  and  day  9  followed  by  a  significant  increase 
between  day  9  and  day  19.  These  responses  resulted  in  a  U-shaped  metabolic 
curve.  Animals  exposed  to  the  fluctuating  temperature  range  of  15-25°  C  showed 
a  significant  metabolic  decrease  (P  <  0.05)  by  day  3  and  another  decrease  (P  < 
0.01)  by  day  6.  The  metabolic  rate  remained  constant  after  this  time  until  day  16 
when  the  experiment  was  discontinued. 

After  exposure  to  three  temperature  ranges  (5-15°  C,  10-20°  C,  and  20-30°  C) 
Panopens  had  the  same  oxygen  consumption  rate  before  and  after  exposure  to  the 
cyclic  thermal  acclimation  period.  In  contrast,  those  crabs  exposed  to  15-25°  C 
had  a  significantly  lower  metabolic  rate  (P  <  0.001)  than  animals  exposed  to  a 
constant  temperature  of  25°  C  (Fig.  2).  Animals  exposed  to  cyclic  temperatures 


MICROLITERS  O2/(hr-g  dry  weight) 

-»•  K) 
O  O 
0  0 

_^l    -t  n       Ort 

U 

C 
C 

c 

1 

Q                                                                            D  CM  5  -25 

•  x 

^ft***"^**^-*       •-!              n                     ««L  .  .« 

^ 
—  »J~I 

tz^>*&?3^~^ 

T—  ^<xa    -""" 

__*--*** 

PANOPEUS  HERBSTII 
I                   1                   1                   1 

1 

CONSTANT     3                  6                  g 

15 

21 

ACCLIMATION  TIME  IN  DAYS 

FIGURE  1.    Oxygen  consumption  of  Pcmopcus  herbstii  after  constant  acclimation  (constant) 
and  after  varying  lengths  of  time  under  the  influence  of  cyclic  temperatures. 


CYCLIC  RESPIRATION  IN  CRABS 


191 


TABLE  II 

Statistical  analysis  of  metabolic-temperature  responses  of  Panopeus  and  Uca  exposed  to  fluctuating 
temperatures  for  different  periods  of  time  (NS  indicates  that  means  are  not  significantly  different). 

Panopeus 
Temperature  range  (°  C) 


Days 

5-15 

10-20 

15-25 

20-30 

0  vs.  3 

NS 

NS 

P  <  0.05 

NS 

3  vs.  6 

NS 

NS 

P  <  0.01 

NS 

6  vs.  9 

NS 

P  <  0.001 

NS 

NS 

9  vs.  last 

NS 

P  <  0.001 

NS 

NS 

Uca 
Temperature  range  (°  C) 


Days 

5-15 

10-20 

15-25 

20-30 

0  vs.  3 

NS 

P  >  0.001 

NS 

NS 

3  vs.  6 

NS 

P  >  0.05 

NS 

NS 

6  vs.  9 

P  >  0.01 

NS 

NS 

NS 

9  vs.  last 

P  >  0.001 

NS 

P  >  0.001 

P  >  0.05 

for  16-19  days  exhibited  an  excellent  degree  of  metabolic-temperature  regulation 
in  that  the  metabolic  rate  was  the  same  at  20°,  25°,  and  30°  C.  In  contrast,  when 
exposed  to  constant  temperature,  the  metabolic-temperature  curve  showed  an  in- 
crease in  metabolism  with  increasing  temperature  until  a  high  stressful  thermal 


O) 

I 

I"     200 
0) 


CM 

o 

tfi 

oc 


100 


o 

oc 
o 


CONSTANT 
CYCLIC 


PANOPEUS  HERBSTII 


15 


20 


25 


30 


TEMPERATURE (  C) 


FIGURE  2.     A  comparison  of  the  oxygen  consumption  of  Panopcus  herbstii  after  constant 
and  cyclic  acclimation  temperatures.    Vertical  brackets  are  the  standard  errors  of  the  means. 


192 


R.  F.  DAME  AND   F.  J.  VERXJ',KK<; 


TAHI.K  III 
values  for  metabolic  rules  of  Panopeus  and  Uca  exposed  to  constant  and  cyclic  thermal  regimes. 


Constant  thermal  regime 

Cyclic  thermal  regime 

Panopeus 

Uca 

Panopeus 

Uca 

15-20 

2.6 

1.5 

2.3 

1.2 

15-25 

2.5 

1.1 

1.6 

1.2 

20-25 

2.3 

1.0 

1.1 

1.2 

20-30 

1.1 

1.2 

1.1 

1.5 

25-30 

<1.0 

1.5 

1.2 

1.9 

point  (30°  C)  was  reached  and  the  rate  decreased.  Ponopcns  hcrbstii  is  killed  at 
temperatures  slightly  over  30°  C.  These  responses  are  better  expressed  as  changes 
in  Qio  values  (Table  III),  in  that  the  Qio  is  less  than  one  over  the  range  of  25- 
30°  C,  but  at  all  other  temperature  ranges  the  Qi0  is  greater  than  one.  The  Qio 
values  of  animals  maintained  at  constant  temperature  are  higher  than  those  of 
animals  subjected  to  a  cyclic  thermal  regime,  except  for  those  at  thermal  ranges 
involving  30°  C.  Of  particular  interest,  animals  exposed  to  a  cyclic  thermal  regime 
did  not  show  the  same  level  of  metabolic  depression  as  did  animals  maintained 
at  constant  temperature.  This  response  could  have  survival  value  to  the  Panopeus 
population. 

The  oxygen  consumption  of  Panopeus  hcrbstii  with  response  to  size  and  oxygen 
concentration  has  been  investigated  by  Leffler  (1973).  This  study  utilized  con- 
stant conditions,  but  a  similar  measurement  technique.  He  showed  that  oxygen 
consumption  was  influenced  by  size  following  the  general  relationship  expressed  by 
von  Bertalanffy  (1957)  and  Hemmingsen  (1960).  Also,  oxygen  consumption  in 
P.  Jicrbstii  dropped  in  rough  proportion  to  the  oxygen  level  of  the  medium.  Neither 
size  nor  oxygen  concentration  factors  should  have  influenced  our  findings  since 
the  crabs  were  of  approximately  the  same  size  (0.5-2.0  grams  dry  weight),  and  the 
crabs  were  partially  exposed  during  the  metabolic  determinations  in  the  Gilson 
respirometer.  The  oxygen  consumption  values  reported  by  Leffler  (1973)  were  of 
the  same  order  of  magnitude  as  those  determined  in  our  study,  but  exact  conversion 
from  Leffler's  to  our  work  is  impossible  since  he  used  a  lower  salinity  (22(/(C}  than 
we  (35/£0).  Recently  Dimock  and  Groves  (1975)  have  shown  that  the  oxygen 
consumption  of  Panopeus  herbstii  is  influenced  by  temperature  and  salinity  com- 
binations. Their  observations  of  oxygen  consumption  are  slightly  lower  than  ours 
at  the  two  most  comparable  temperature  and  salinity  combinations  (10°  C  and 
30%0;  25°  C  and  30^c).  The  reduced  metabolic  vaiues  of  Dimock  and  Groves' 
data  are  probably  the  result  of  using  a  full  range  of  size  classes,  thus  a  larger 
average  size  being  used  in  their  studies. 

The  respiratory  responses  of  Uca  puijilator  to  increasing  cyclic  temperature  ac- 
climation time  were  different  than  those  of  Panopeus  and  were  highly  variable 
(Fig.  3).  Interspecific  differences  were  noted  in  that  the  fiddler  crabs  had  the 
higher  metabolic  rate  at  each  temperature.  This  result  is  consistent  with  the 
earlier  finding  of  Vernberg  (1969)  that  crabs  which  exhibit  a  high  level  of  locomo- 


CYCLIC  RESPIRATION  IN  CRABS 


400r 


•O      3OO 

o> 


£      200 

t 

O 
O 

I 

100 


-O   5  -  15  C 

-•  10  -  20  c 

Q  15  -  25  C 
2O  -  30  C 


UCA  PUGILATOR 


I 


CONSTANT     3 


69  15 

ACCLIMATION  TIME  IN  DAYS 


22 


FIGURE  3.  Oxygen  consumption  of  Uca  pugilator  after  constant  acclimation  (constant)  and 
after  varying  lengths  of  time  under  the  influence  of  cyclic  temperatures. 

tor  activity  tend  to  have  a  higher  metabolic  rate  than  a  more  lethargic  species,  Uca 
is  an  active  crab  at  low  tide  when  it  can  be  seen  darting  about,  while  Panopcus  is 
more  secretive,  hiding  among  the  oyster  shells.  Also,  some  specimens  of  Uca  are 
more  temperature  tolerant  than  Panopcus;  an  exposure  to  30°  C  is  less  stressful 
on  their  metabolic  response. 

Following  different  exposure  times  to  fluctuating  temperatures,  significant  dif- 
ferences in  means  of  metabolic  rate  were  observed  at  each  thermal  regime  (Table 
II).  In  contrast,  significant  differences  occurred  at  only  two  of  the  four  tempera- 
ture ranges  for  Panopcus.  However,  when  comparing  the  initial  metabolic  rate 
with  the  rate  at  the  end  of  the  exposure  to  fluctuating  temperatures,  only  at  20°  C 
was  there  a  statistically  significant  difference  with  the  rate  being  lower.  After  ex- 
tended cyclic  temperature  acclimation,  specimens  of  I'ca  acclimated  to  the  warmest 
range  (20-30°  C)  consumed  significantly  more  oxygen  than  the  other  three  groups 
which  had  similar  rates  (significant  at  the  95^  confidence  level).  This  trend  sug- 
gests that  specimens  of  Uca  are  less  sensitive  to  thermal  change  than  are  those  of 
Panopcus.  Further  evidence  for  this  statement  is  that  the  Q1()  values  for  Uca  tend 
to  be  lower  than  those  for  Panopcus  (Table  III). 

A  graphic  representation  of  the  oxygen  consumption  rates  of  crabs  maintained 
at  constant  and  cyclic  temperature  is  shown  in  Figure  4.  The  metabolic  rate  of 
Uca  pugilator  showed  significant  differences  between  constant-  and  cyclic-tem- 
perature-acclimated crabs  at  20°  C  (95r/r  level),  25°  C  (92%  level),  and  15°  C 
(93 %  level),  but  not  at  the  other  temperature  of  30°  C.  The  metabolic  response  of 
Panopcus  varied  significantly  only  at  25°  C.  In  the  case  of  both  species,  the  oxygen 
consumption  rate  was  lower  in  animals  subjected  to  cyclic  temperatures. 

The  metabolic  responses  of  Uca  pugilator  have  been  extensively  investigated 
by  Vernberg  (1969).  Constant-temperature-acclimated  specimens  of  U.  pugilator 
from  the  present  study  exhibited  the  same  acclimation  curve  and  metabolic  rates  as 


194 


R.  F.  DAME  AND   F.  J.  VERNBERG 


~.    4OO 


•o 

01 


30° 


N 

O 

(/) 
DC 
U 

j 
O 

cc    200 
o 


CONSTANT 
CYCLIC 


UCA  PUGILATOR 


15 


20 


25 


3O 


TEMPERATURE (  C) 


FIGURE  4.     A  comparison  of  the  oxygen  consumption  of  Uca  pugilator  after  constant  and 
cyclic  acclimation  temperatures.    Vertical  brackets  are  the  standard  errors  of  the  means. 

crabs  from  North  Carolina.  The  cyclic-temperature-acclimated  specimens  of  U. 
pugilator  showed  a  depressed  response  compared  to  that  of  constant-temperature- 
acclimated  crabs. 


DISCUSSION 

The  experiments  described  here  offer  evidence  that  the  metabolic  rate  of  inter- 
tidal  organisms,  as  measured  by  oxygen  consumption,  is  influenced  differently  by 
constant  and  cyclic  temperature  acclimation  regimes.  The  evidence  is  most  striking 
in  Panopeus  exposed  to  the  15-25°  C  cycle  and  Uca  exposed  to  5-15°,  10-20°,  and 
15-25°  C  cycles.  Other  workers  have  demonstrated  differences  in  various  physio- 
logical response  that  could  be  correlated  with  variations  in  the  type  of  thermal  ac- 
climation regime. 

Thorp  and  Hoss  (1975)  determined  that  cyclic  temperatures  decreased  survival 
of  the  grass  shrimp,  Palaemonetes  pit(/io  and  P.  rulyaris,  at  low  salinities  (5%) 
and  low  cyclic  temperatures  (7-13°  C)  when  compared  to  shrimp  kept  at  constant 
temperatures  under  the  same  salinity.  In  contrast,  the  pupfish,  Cyprinodon 
ncvadcnsis,  acclimated  to  cycling  temperatures  demonstrated  a  greater  tolerance  to 
both  high  and  low  temperature  than  animals  acclimated  to  a  constant  temperature 
(Feldmeth,  Stone,  and  Brown,  1974).  This  increased  scope  of  thermal  tolerance 
could  have  survival  value  to  animals  occupying  habitats  which  characteristically 
are  subjected  to  wide  daily  and  seasonal  thermal  changes. 

Cyclic  temperatures  may  influence  survival  and  development  of  larvae.  Costlow 
and  Bookhout  (1971)  found  that  survival  of  the  larval  mud  crab,  Rhithropanopeus 
harrisn,  was  about  the  same  for  a  cycled  temperature  as  for  a  constant  temperature 
equal  to  the  mean  of  the  cycled  temperatures  over  the  range  of  10-30°  C,  but 


CYCLIC  RESPIRATION  IN  CRABS  195 

larvae  maintained  at  a  cycle  of  30-35°  C  survived  better  than  those  at  either  30°  or 
35°  C.  The  time  required  to  complete  metamorphosis  at  the  warmer  cycles  was 
influenced  by  the  high  temperature  in  the  cycle  at  all  salinities  tested  (Christiansen 
and  Costlow,  1975).  However,  the  results  of  these  two  studies  at  higher  tem- 
peratures are  different,  possibly  the  result  of  variation  in  survival  of  crab  larvae 
from  hatch  to  hatch.  More  recently,  Sastry  and  Yargo  (1977)  found  that  larvae 
of  a  decapod  crustacean.  Cancer  irroratits,  showed  a  greater  survival  rate  when 
reared  under  a  suitable  amplitude  and  rate  of  temperature  change  than  larvae 
maintained  at  comparable  constant  temperatures. 

Cyclic  temperature  may  influence  sublethal  responses  of  organisms,  but  not 
always  in  an  apparent,  predictable  way.  Hoffman  (1974),  working  with  crickets, 
reported  that  varying  day-night  temperature  cycles  do  not  accelerate  physiological 
functions,  except  for  life  span  and  egg  production. 

Thorp  and  Hoss  (1975)  determined  the  oxygen  consumption  rate  of  two  species 
of  grass  shrimp  after  acclimation  to  constant  temperatures  (7°  and  10°  C)  and 
cyclic  temperatures  (7°-13°  C)  at  salinities  of  5,  20.  and  35/£o.  Both  species  of 
shrimp  utilized  more  oxygen  when  acclimated  to  constant  temperature  (10°  C) 
and  35',,  than  cyclic  temperatures  of  7-13°  C  and  35/u.  This  result  is  similar  to 
our  work  in  that  our  crabs  were  acclimated  to  35/<r,  and  when  differences  in  oxygen 
consumption  were  observed,  cyclic  temperatures  depressed  oxygen  consumption. 

Unlike  our  results,  the  respiration  rate  of  spiders  maintained  on  a  cyclic  tem- 
perature regime  was  higher  than  animals  kept  at  a  constant  temperature  (Hum- 
phreys, 1975).  This  response  may  be  correlated  with  an  increase  in  growth  and 
development  rates  of  this  species.  However,  since  the  resting  metabolic  rate  is 
higher  when  kept  on  cyclic  thermal  regimes,  the  organisms  must  be  more  efficient 
in  extracting  energy  from  their  food  or  they  must  eat  more  in  order  to  grow. 
Mytilus  edulis,  a  bivalve,  utilizes  a  different  strategy  in  adapting  to  cyclic  tempera- 
tures. It  reduces  the  amplitude  of  both  oxygen  consumption  and  filtration  rate 
(Widdows,  1976).  Specimens  of  Panopcus  and  Uca  respond  metabolically  in  a 
similar  manner,  at  least  at  intermediate  temperatures.  One  result  of  reducing  the 
standard  metabolic  rate  is  to  conserve  energy  which  can  be  used  for  other  functional 
activities  necessary  for  an  organism  to  successfully  compete  and  survive.  This  is 
particularly  important  for  these  crabs  in  that  the  cyclic  thermal  ranges  at  which 
they  demonstrate  reduced  rates  of  oxygen  uptake  are  those  in  which  these  organisms 
are  most  active  during  most  of  the  year. 

Based  on  the  results  of  this  study  it  seems  that  any  previous  estimate  of  the 
role  of  oxygen  consumption  in  energy  budgets  of  a  species  and/or  a  community 
might  be  in  error,  unless  the  influence  of  cyclic  temperature  on  respiration  has  been 
determined. 


"We  would  like  to  thank  Bill  Murtiashaw  for  his  valuable  assistance  in  the  meta- 
bolic measurements  and  acclimation  procedures.  Stuart  Stevens  and  Bill  Johnson 
provided  data  processing  assistance. 


!')()  K.  F.  DAME  AND  F.  J.  VERNBERG 

SUMMARY 

The  comparative  influence  of  acclimation  to  constant  and  cyclic  temperatures  on 
the  metabolic  rates  of  the  mud  crab,  Panopcits  licrhstii,  and  the  fiddler  crab,  Uca 
pugilator,  was  observed.  Although  interspecific  differences  were  observed,  cyclic 
acclimation  temperatures  significantly  depressed  oxygen  consumption  in  the  15°- 
25°  C  temperature  range  in  both  species  when  compared  to  rates  of  animals  sub- 
jected to  constant  acclimation  rates.  Since  this  depression  of  metabolic  rates  occurs 
over  that  portion  of  the  yearly  temperature  range  within  which  the  animals  are  most 
active,  it  is  suggested  that  these  organisms  utilize  energy  more  efficiently  when 
subjected  to  natural  cyclic  temperature  conditions  than  when  subjected  to  constant 
temperature  environments.  This  difference  in  metabolic  data  would  be  significant 
in  analyzing  the  role  of  the  yearly  energy  budgets  of  crabs  in  ecosystem  energetics. 

LITERATURE  CITED 

BAYNE,  B.  L.,  J.  WIDDOWS,  AND  C.  WORRALL,  1977.  Some  temperature  relationships  on  the 
physiology  of  two  ecologically  distinct  bivalve  populations.  Pages  379-400  in  F.  J. 
Vernberg,  A.  Calabrese,  F.  Thurberg,  and  W.  Vernberg,  Eds.,  Physiological  responses 
of  marine  biota  to  pollutants.  Academic  Press,  New  York. 

CHRISTIANSEN,  M.  C.,  AND  J.  D.  COSTLOW,  1975.  The  effect  of  salinity  and  cyclic  temperature 
on  larval  development  of  the  mud-crab  Rhithropanopeus  harrisii  reared  in  the  labora- 
tory. Mar.  Biol.,32:  215-221. 

COSTLOW,  J.  D.,  AND  C.  G.  BOOKHOUT,  1971.  The  effect  of  cyclic  temperatures  on  larval 
development  in  the  mud-crab  Rhithropanopeus  Iiarrisii.  Pages  211-220  in  D.  J.  Crisp, 
Ed.,  Fourth  European  iiiuriiie  biology  symposium.  Cambridge  University  Press,  Eng- 
land. 

DEHNEL,  P.,  1958.  Effect  of  photoperiod  on  the  oxygen  consumption  of  two  species  of  inter- 
tidal  crabs.  Nature.  181:  1415-1417. 

DIMOCK,  R.  V.,  AND  K.  H.  GROVES,  1975.  Interaction  of  temperature  and  salinity  on  oxygen 
consumption  of  the  estuarine  crab  Panopeus  lierhstii.  Mar.  Biol.,  33:  301-308. 

FELDMETH,  C.  R.,  E.  A.  STONE,  AND  J.  H.  BROWN,  1974.  An  increased  scope  for  thermal 
tolerance  upon  acclimating  pupfish  (Cyprinodon)  to  cycling  temperatures.  /.  Com  p. 
Physiol.,  89:  39-44. 

HEMMINGSEN,  A.  M.,  1960.  Energy  metabolism  as  related  to  body  size  and  respiratory  surfaces 
and  its  evolution.  Rep.  Steno  Mem.  Hasp.,  9:  1-110. 

HOFFMAN,  K-H.,  1974.  Wirkung  von  konstanten  und  tagesperiodish  alternierenden  tempera- 
turen  auf  lebensdauer,  nahrungsverwertung  und  fertilitat  adulter  Gr\llus  bimaculatus. 
Occoloyia,  17 :  39-54. 

HUMPHERYS,  W.  F.,  1975.  The  respiration  of  Gcolycosa  godcffroyi  (Araneae:  Lycosidae) 
under  conditions  of  constant  and  cylic  temperature.  Physiol.  Zoo/.,  48:  269-281. 

KAHN,  M.  F.,  1965.  The  effect  of  constant  and  varying  temperatures  on  the  development  of 
Acanthocyclops  viridis.  Proc.  R.  Ir.  Acad.  Sect.  B,  64:  117-130. 

LEFFLER,  C.  W.,  1973.  Metabolic  rate  in  relation  to  body  size  and  environmental  oxygen  con- 
centration in  two  species  of  xanthid  crabs.  Camp.  Bioehein.  Physiol.,  44A  :  1047-1052. 

NEWELL,  R.  C.,  1975.  Factors  controlling  metabolic  capacity  adaption  in  marine  invertebrates. 
Pages  111-128  in  F.  J.  Vernberg,  Ed.,  Physiological  ecology  of  estuarine  organisms. 
University  of  South  Carolina  Press,  South  Carolina. 

SASTRY,  A.  N.,  AND  S.  L.  VARGO,  1977.  Variations  in  the  physiological  responses  of  crustacean 
larvae  to  temperature.  Pages  401-424  in  F.  J.  Vernberg,  A.  Calabrese,  F.  Thurberg, 
and  W.  Vernberg,  Eds.,  Physiological  responses  of  marine  biota  to  pollutants.  Aca- 
demic Press,  New  York. 

STEEL,  R.  G.  D.,  AND  J.  H.  TORKIE,  I960.  Principles  ami  procedures  in  statistics.  McGraw- 
Hill,  New  York. 


CYCLIC  RESPIRATION  IN  CRABS  197 

THORP,  J.  H.,  A  xi>  D.  E.  Hoss,  1975.  Effects  of  salinity  and  cyclic  temperatures  on  survival 
of  two  sympatric  species  of  grass  shrimp  and  their  relationship  to  natural  distributions. 
J.  E.rp.  'Mar.  Biol.  Ecol.,  18:  19-28. 

VAX  \YIXKLK,  \Y.,  1969.  Physiological  effects  of  short  term  cyclic  environmental  changes. 
Am.  Zoo}.,  9:  1100. 

VERNBERG,  F.  J.,  1969.    Acclimation  of  intertidal  crahs.    Am.  Zonl..  9  :  333-341. 

YEKNBERG,  F.  J.,  AXIJ  W.  B.  VERNBERG,  1966.  Studies  on  physiological  variation  between  tropi- 
cal and  temperate  zone  fiddler  crabs  of  the  genus  Uca.  VII.  Metabolic-temperature  ac- 
climation responses  in  southern  hemisphere  crabs.  Comp.  Biochciu.  Physiol.,  19:  489- 
524. 

VERNBERG,  \V.  B.,  AND  F.  J.  VEKNBERG,  1972.  Environmental  physiology  of  marine  animals. 
Springer- Verlag,  New  York,  Heidelberg,  and  Berlin. 

vox  BERTALANFFY,  L.,  1957.  Quantitative  laws  in  metabolism  and  growth.  Q.  Rev.  Biol.,  32: 
237-241. 

WIDDOWS,  J.,  1976.  Physiological  adaptation  of  Mvtilus  cdulis  to  cyclic  temperatures.  /.  Comp. 
Physiol.,  105:  115-128. 


Refavm-e:  Bwl.  Bull.,  154:    1W-212.      (April,   1978) 


MULTIPLE  NUCLEI  DURING  EARLY  OOGENESIS  IN  FLECTONOTUS 
PYGMAEUS  AND  OTHER  MARSUPIAL  FROGS 

EUGENIA  M.  DHL  PINO  AND  A.  A.  HUMPHRIES,  J  R. 

Inslitiilo  (/<'  Cicadas,  Pontificia  Universidad  Catolica  del  Ecuador,  Quito,  Ecuador;  and 
Department  of  Biology,  Emory  University,  Atlanta,  Georgia  30322 

In  the  great  majority  of  amphibians  yet  investigated  the  occurrence  of  oocytes 
with  more  than  one  nucleus  or  more  than  a  single  meiotic  figure  is  exceptional 
(Humphries,  1956,  1966;  Parmenter,  Derezin  and  Parmenter,  1960;  Humphrey, 
1963).  In  the  tailed  frog  Ascaplins  tniei,  however,  Macgregor  and  Kezer  (1970) 
found  that  oogenesis  regularly  involves  oocytes  with  eight  nuclei,  all  hut  one  of 
which  disappear  before  the  final  stages  of  oogenesis.  When  a  related  species, 
Leiopelma  hochstetteri,  was  investigated  by  Robinson,  Stephenson  and  Stephenson 
(1973),  only  a  single  binucleate  oocyte  was  found  among  26  oocytes  examined.  In 
the  present  paper  the  occurrence  of  multinncleate  stages  as  a  regular  feature  of 
oogenesis  in  several  genera  of  marsupial  frogs  of  South  America  is  reported.  The 
multinucleate  condition  is  associated  with  the  early  stages  of  oogenesis ;  in  large, 
yolky  oocytes  only  one  nucleus  is  present. 

Most  of  the  marsupial  frogs  are  inhabitants  of  the  humid  forests  of  South 
America.  In  these  frogs  the  aquatic  larval  stages  are  reduced  or  eliminated  al- 
together, a  phenomenon  associated  with  the  fact  that  the  female  carries  the  em- 
bryos on  her  back,  either  within  a  pouch  of  integument  or  in  shallow  depressions 
of  the  skin.  The  genera  Flectonotus,  Gas/rotheca,  and  Amphignathodon  are  char- 
acterized by  pouches,  while  in  Fritziana,  Cryptobatrachus,  Stefania  and  Hcini- 
pJiractiis  the  embryos  are  carried  in  depressions  of  the  skin.  For  a  list  of  these 
frogs  with  localitv  and  references  of  taxonomic  interest  see  the  work  of  Duellman 
(1976). 

Relatively  little  information  is  available  regarding  reproduction  and  development 
in  marsupial  frogs.  The  relationship  between  mother  and  embryos  has  been 
studied  in  some  detail  in  Gastrotheca  riobauibae,  a  species  that  carries  the  embryos 
in  the  pouch  up  to  the  tadpole  stage  (see  del  Pino,  Galarza,  de  Albuja,  and  Hum- 
phries, 1975,  for  references).  Among  the  species  of  Gastrotheca  that  carry  the 
embryos  to  the  young  froglet  stage,  Gastrotheca  oi'ifcra  is  the  best  known  (see 
Mertens,  1957,  for  a  description  of  its  life  history  and  references). 

This  report  describes  some  features  of  oogenesis  in  33  species  of  marsupial 
frogs  that  correspond  to  the  described  species  of  Amphignathodon  and  Heini- 
[>hraclns  as  listed  by  Duellman  (1976),  to  two  (of  three)  species  of  Flectonotns, 
one  (of  six)  species  of  Stefania  and  20  (of  32)  species  of  Gastrotheca.  In  addi- 
tion, the  ovaries  of  three  unnamed  species  of  Gastrotlieca  were  also  analyzed.  The 
peculiarities  of  oogenesis  have  been  studied  in  more  detail  in  Flectonotus  pygiiiaeus, 
Gastrotheca  orijera  and  Gastrotlieca  sp.,  an  unnamed  species  from  Venezuela,  since 
in  these  instances  both  living  and  preserved  specimens  were  available.  Living 
specimens  of  G.  marsupiata,  G.  c.vcubitor,  G.  nierteusi  and  G.  plumbea  were  also 

198 


MULTINUCLEATE  OOGEXKSIS 


available,  but  only  in  limited  numbers.  Study  ot  other  frogs  was  restricted  to 
museum  specimens,  \vitb  tbe  exception  of  G.  riobambae,  in  which  the  ovary  had 
been  studied  previously  (del  Pino  and  Sanchez,  1977). 

MATERIALS  AND  METHODS 
Specimens  examined  and  laboratory  care  of  living  frogs 

Flectonohts  f>ygntaetis,  Gastrolhcca  orifera  and  Gastrotheca  sp.  were  collected 
at  Estacion  Biologica  de  Rancho  Grande,  Maracay,  Estado  de  Aragua,  Venezuela 
in  November  1975  and  transported"  alive  to  the  laboratory  in  Quito.  In  addition, 
several  specimens  preserved  in  W/c  formalin  or  in  Bouin's  picro-formol  were 
available  for  study.  Flectonotits  pyymacits  gives  birth  to  advanced  tadpoles,  while 
G.  ovijera  and  Gastrotheca  sp.  give  birth  to  froglets.  For  this  analysis  the  ovaries 

TABLE  I 
Size  of  ovarian  oocytes  of  Flectonotus  pygmaeus  at  various  reproductive  stages. 


Reproductive  stage 

Oocyte 
diameter 

<Mm) 

Number  of  large  oocytes 

Number  of 
embryos  in 
the  pouch 

Left  ovary 

Right  ovary 

Total 

Juvenile 

Without  pouch 

500 

4 

7 

11 

500 

(  )p(.'n  pouch 

1200 

0 

5 

5 

1500 

2000 

3 

3 

6 

2500 

3 

6 

9 

Adult 

Closed  pouch 

3000 

4 

9 

13 

3000 

6 

5 

11 

3000 

5 

3 

8 

Pregnant 

4  days 

800 

3 

3 

6 

6 

8  days 

1000 

6 

7 

13 

8 

18  days 

500 

6 

20  days* 

1500 

3 

2 

5 

7 

20  days 

2000 

4 

5 

9 

8 

23  days 

2000 

6 

6 

12 

11 

28  days 

500 

7 

At  birth 

2000 

1 

11 

12 

11 

2000 

1 

5 

6 

5 

2000 

1 

4 

5 

8 

2500 

T 

3 

5 

8 

7 

2000 

4 

5 

9 

6 

After  birth 

2  days 

1500 

6 

2 

8 

7 

3  days 

500 

8 

4  days 

1500 

2 

3 

5 

6 

Mean  =fc  s.d. 

3.5  ±  1.8 

5.0  ±  2.3 

8.5  ±  2.8 

7.4  ±  1.7 

Frog  was  killed  20  days  after  ovulation;  mating  did  not  occur. 


2()()  E.  M.   IIKI.  P1XO   AND   A.   A.    1 1 1 '  M  IM  I  K  I  KS 

of  25  1' .  f>v</niiifits  females  (Table  I  )  and  llio.se  ot   live  (/'.  ovifera  and  two  (,'iislro- 
Ihcca  sp.  were  studied. 

Juvenile  specimens  of  /;.  pyi/inucus  were  kept  in  a  humid  terrarium  at  20°  C 
(range  17-21°  C)  and  were  fed  DrosopJiila  once  or  twice  a  day.  The  frogs  lived 
for  nearly  a  year.  Reproductive  activity  occurred  spontaneously  after  about  five 
months  of  captivity.  Three  females  became  pregnant;  a  fourth  deposited  eggs  but 
did  not  mate.  In  the  latter  case,  the  eggs  were  deposited  on  the  wall  of  the  ter- 
rarium. One  of  these  eggs  was  fixed  with  Houin's  picro-formol  for  cytological 
examination.  Under  laboratory  conditions,  the  period  of  incubation  in  the  pouch 
lasts  about  29  days ;  advanced  tadpoles  then  emerge  from  the  pouch  and  metamor- 
phose after  about  30  days.  Incubation  in  the  pouch  of  this  species  is  considerably 
abbreviated  in  comparison  with  the  situation  in  (/".  rlobainlntc,  where  development 
in  the  pouch  lasts  nearly  four  months  (del  Pino  ct  al.,  1975). 

Most  of  the  museum  specimens  that  were  examined  belong  to  the  Museum  of 
Natural  History  of  the  University  of  Kansas  (KU).  Other  specimens  belong 
to  The  American  Museum  of  Natural  History  (AMNH),  Museo  de  la  Estacion 
Biologica  de  Rancho  Grande  (EBRG),  the  Field  Museum  of  Natural  History 
(FMNH),  Museo  de  la  Fundacion  Miguel  Lillo  (MFML),  the  Museum  of  Zool- 
ogy of  the  Louisiana  State  University  (LSUMZ)  or  the  Museum  of  Natural 
History  of  the  University  of  Southern  California  (USC).  Information  regarding 
the  museum  specimens  is  presented  below. 

AMPHIGNATHODON:    A.  t/iicnthcri  KU-164228,  Ecuador,  Pichincha :  5  km  ESE  Chiriboga, 
2010  m. 

CRYPTOBATRACHUS:     C.  fuhnnnnni  KU-169378,  Colombia,  Norte  de  Santander,  32  km 
W  Sardinata,  1050  m. 

FLECTONOTUS  :    F.  fissilis  KU-92240,  Brazil,  Guanabara  :  Rio  de  Janeiro,  Tijuca. 

GASTROTHECA:  G.  ari/cntcorircus  KU-144123,  Colombia,  Cauca :  Road  to  Pacific  coast 
from  El  Tambo,  2170  m.  G.  caria  KU-148534,  Ecuador,  Imbabura:  Laguna  de  Cuicocha, 
2890  m.  G.  ceratophrys  KU-77016,  Panama,  Darien,  Laguna,  820  m. ;  AMNH-90984. 
G.  christiani  MFML-02117,  MFML-02117-5,  Argentina,  Jujuy,  Abra  de  Canas  V.  Grande. 
G.  cornuta  KU-169394,  Colombia,  Cauca:  La  Costa,  El  Tambo,  1000  m.  G.  c.rcubitor  KU- 
163135,  KU-163140,  Peru,  Abra  Acanacu,  25  km  NNE  Paucartambo,  3520  m.  G.  gracilis 
MFML-01972,  Argentina,  Tucuman,  Road  Tafi  del  Valle  km  41.  G.  grisivoldi  KU-138221, 
Peru,  Junin.  Mayupampa,  21  km  N  La  Oroya,  3400  m;  KU-138227,  Peru,  Junin,  Comas 
3220  m.  G.  lojana  KU-138234,  Ecuador,  Loja,  10  km  W  Loja  2500  m;  KU-142608,  Ecua- 
dor, Loja,  5.5  km  W  Loja,  2330  m.  G.  marsnpiata  KU-138399,  Peru,  Huancavelica :  Huan- 
cavelica  3780  m;  KU-139187,  Peru,  Cuzco,  14.5'  km  S  Paucartambo,  3450  m.  G.  incrtcusi 
KU-140386,  Colombia,  Narino:  La  Victoria,  2700  m.  G.  microdisca  KU-154610,  Brazil, 
Sao  Paulo:  10  km  NW  Caraguatatuba  500-750  m.  G.  manticola  KU-138402,  Ecuador, 
Azuay:  Giron  2240-2500  m;  KU-142610,  Eucador,  Loja:  Saraguro,  2510  m.  G.  ochoai 
KU-138668,  Peru,  Puno :  Ollachea,  53  km  W.  Macusani,  2800  m.  G.  ovifera  KU- 
125372,  Venezuela,  Distrito  Federal  de  Caracas;  KU-13338,  Venezuela,  Aragua :  Quebrada, 
0.5  km  E.  Res.  Sta.  at  Rancho  Grande,  1075  m.  G.  pcniana  KU-138444,  Peru,  Huanco 
5  km  NE  La  Union  3100  m;  KU-138494,  Peru,  Cajamarca:  Cajamarca,  2800  m;  KU- 
138526,  Peru,  Ancash,  Chavin  de  Huantar  3230  m.  G.  plumbed  KU-132414,  Ecuador, 
Cotopaxi:  Pilalo  2460-2580  m;  KU-164230,  Ecuador,  Pichincha:  9.5'  km  NW  Nono  2530 
m.  G.  tcstiidinca  KU-163276,  Peru,  Ayacucho :  Tutumbaro,  Rio  Piene,  1840  in.  G.  u>cin- 
landii  KU-146042,  Ecuador,  Morona  Santiago:  Rio  Piuntza,  1830  m.  Gastrothcca  sp., 
FMNH-39889.  Gastrothcca  sp.,  LSUMZ-32049.  Gastrothcca  sp.,  EBRG-48240,  Venezuela, 
Aragua :  Res.  Sta.  at  Rancho  Grande. 

HEMIPHRACTUS:    H.  bubalus  KU- 169426,  Colombia,  Putumayo:    10.3  km  W  El  Pepino, 
1440  m.     H.  fasciatus  KU-93503,   Panama,   Altos  de   Pacora;   KU-1 16353,    Panama,    San 


MULTINUCLEATE  OOGENESIS  201 

Bias:  Camp  Summit,  300-400  m.  H.  johnsoni  USC-716,  Peru.  II.  proboscideus  KU- 
123139,  Ecuador,  Napo :  Santa  Cecilia,  340  m.  //.  scnfutns  KU-147118,  Ecuador,  Morona 
Santiago:  Rio  Piiinza,  1400m. 

STEFANIA:  S.  scalar  KU-167222.  Venezuela,  Bolivar:  Paso  de  El  Danto,  El  Dorado-Sta. 
Elena  de  Vairen  road,  km  117-119  (100-115(1  m)  ;  KU-16239,  KU-167248,  Venezuela, 
Bolivar:  El  Dorado-Sta.  Elena  de  Vairen  road,  km  112,  860  m. 

Cytological  procedures 

The  nuclei  of  ovarian  oocytes  were  observed  in  both  living  and  histological 
preparations.  Oocytes  from  100-500  p.m  are  somewhat  transparent  and  can  be 
studied  intact.  Observations  were  made  on  such  oocytes  using  a  depression  slide 
or  a  standard  slide  on  which  the  coverglass  was  slightly  elevated.  Individual  nuclei 
were  observed  similarly  with  phase  contrast  microscopy  after  rupturing  the  oocyte 
in  Ringer's  solution  or  a  5  :  1  mixture  of  0.1  M  XaCl  and  0.1  M  KC1. 

Ovarian  tissue  fixed  in  10r-f  formalin  or  in  Bouin's  picro-formol,  as  well  as  that 
from  museum  material,  was  embedded  and  cut  into  sections  of  10  ^m  thickness  and 
stained  with  Harris'  hematoxylin.  Alcoholic  eosin  yellow  was  used  as  the  counter- 
stain.  Some  preparations  were  made  using  the  standard  Feulgen  procedure. 

The  number  of  nuclei  per  oocyte  was  estimated  by  counts  in  the  sectioned  ma- 
terial. The  section  interval  for  the  counts  was  decided  by  measurements  of  nuclear 
diameter.  Since  the  size  of  nuclei  within  a  single  oocyte  is  highly  variable  at  some 
stages,  the  estimates  of  total  nuclear  number  are  subject  to  considerable  error. 
Sections  of  museum  material  were  unsuitable  for  nuclear  counts,  thus  only  rough 
appraisals  were  made  in  those  cases. 

Ovaries  of  two  specimens  of  I7.  py(/inacits  were  labeled  with  3H-uridine.  Each 
frog  received  an  intracoelomic  injection  of  40  /iCi  of  3H-uridine  (New  England 
Nuclear,  specific  activity  26.7  Ci/mmole)  in  amphibian  Ringer's  solution  and  was 
sacrificed  24  hours  later.  Pieces  of  ovary  were  fixed  in  Bouin's  picro-formol  or  in 
3%  glutaralclehyde  in  a  phosphate  buffer  at  pH  7.4.  The  tissues  were  embedded 
in  Paraplast  and  cut  into  sections  of  10  //.m  thickness.  After  incubation  in  10% 
trichloracetic  acid  at  room  temperature  for  one  hour  the  slides  were  coated  with 
photographic  emulsion  (Ilford  Nuclear  Research  emulsion  type  K-2.  in  gel  form) 
and  stored  for  20  days  at  4°  C.  Following  development,  the  slides  were  stained 
with  Harris'  hematoxylin  and  alcoholic  eosin  yellow. 

RESULTS 
Oogencsis  in  Flectonotus  pygmaeus 

The  ovary  of  F.  pyr/macns  produces  relatively  few  mature  eggs  per  breeding 
period  (Table  I),  but  each  egg  accumulates  a  considerable  store  of  yolk  and  reaches 
a  final  diameter  of  about  3  mm.  The  right  ovary  generally  contains  more  large 
oocytes  than  the  left  (Table  I),  and  sometimes  the  left  ovary  appears  to  be  absent 
altogether.  There  are,  however,  individual  differences,  and  in  some  frogs  the  left 
ovary  is  the  larger  one. 

In  the  mature  ovary  there  are  large,  yellowish-white  yolky  oocytes  of  1500  to 
3000  JU.ITI  diameter,  a  number  of  previtellogenic  oocytes  of  about  500  fim  diameter, 
and  smaller  oocytes  and  oogonia.  Following  ovulation,  oocytes  pass  through  the 


202 


E.  M.  DEL  PINO  AND  A.  A.  HUMPHRIES 


. 


V," 


FIGURE  1.  Cross  section  of  the  ovary  of  Flcctonotus  pyoinacus.  The  two  centrally  located 
oocytes  show  nuclei  of  various  sizes,  with  the  larger  nuclei  distributed  towards  the  periphery 
and  the  smaller  toward  the  center  of  the  oocyte.  In  the  oocyte  on  the  left,  most  nuclei  have 


MULT1NUCLEATE  OOGENESIS  203 

oviducts  and  become  covered  with  a  thin  coat  of  jc-lly  similar  to  that  in  Gastrotheca 
riobauibac  (del  Pino  ct  <//.,  1()75  i.  During  amplexus  the  eggs  are  fertilized  as  they 
leave  the  cloaca  and  are  then  moved  into  the  pouch;  however,  the  details  of  the 
process  are  not  yet  known.  The  oocytes  of  intermediate  size  grow  rapidly  in  the 
mother's  ovaries  during  the  period  that  embryos  are  being  incubated  in  the  pouch 
and  hv  the  time  of  tadpole  birth  these  oocytes  are  vitellogenic  and  about  2  mm  in 
diameter  (Table  1  ).  Only  after  birth  of  the  tadpoles  do  the  oocytes  attain  their 
mature  size  of  about  3  mm. 

The  time  required  for  the  growth  of  oocytes  from  the  initial  .stages  to  ovula- 
tion  is  not  known.  \  itellogenesis,  however,  seems  to  occur  in  a  matter  of  a  few 
months,  since  the  oocytes  of  frogs  kept  in  captivity  grew  in  about  five  mouths  from 
500 -X(K)  /iin  diameter  to  their  mature  size.  It  seems  likely  that  in  nature  the  period 
of  ovarian  growth  might  be  shorter. 

(  )ogonial  divisions  and  various  stages  of  oogenesis  are  easily  observed  in  the 
adult  ovary  (  Figs.  1,2).  Small  oocytes  contain  from  about  1000  to  3000  nuclei 
(Table  II  )  ;  in  larger  oocytes  the  number  of  nuclei  decreases  gradually  (Figs.  8,  9) 
until,  at  later  stages,  only  one  nucleus  remains  (Figs.  10,  11).  Early  oogenesis  oc- 
curs within  chambers  referred  to  as  cysts,  following  the  terminology  of  King  (  1(>US). 
Oogonial  mitoses  within  cysts  appear  to  be  synchronous,  but  division  of  the  cyto- 
plasm does  not  always  accompany  nuclear  division  ;  thus  occasional  multinucleate 
oogonia  are  observed.  Fusion  of  the  cells  of  a  cyst  appears  to  produce  the  multi- 
nucleate  oocyte  (Fig.  2).  Assuming  that  all  the  nuclei  in  a  given  cyst  are  the  division 
products  of  an  original  oogonia]  cell,  there  are  about  11  rounds  ot  milosis  involved 
in  the  formation  of  the  multinucleate  aggregate  that  becmiu's  the  oocyte.  The  ag- 
gregate measures  about  100  /mi  in  diameter  and  contains  some  2000  nuclei  (Table 
II).  There  are  large  differences  in  the  number  of  nuclei  from  one  aggregate  to 
another,  but  the  size  of  the  nuclei  is  generally  uniform  ;  however,  there  are  a  few 
exceptionally  large  nuclei  which  may  lie  polyploid  (Fig.  2). 


disappeared,  and  the  cytoplasm  is  nut  homogeneous.  Cysts  (arrows)  are  found  Irequenlly. 
Bar  represents  100  /j.m. 

FIGURE  2.  Cross  section  of  a  cyst  in  the  ovary  of  F.  />v.<//;/</r;/.v.  Tin-  size  of  the  nuclei  is 
mostly  uniform,  but  an  occasional  large  nucleus  (arrow)  is  seen.  Cell  membranes  were  present 
between  the  oogonia  of  the  cyst,  but  these  seem  to  disappear  as  the  oocyte  forms.  Jn  I  lie  oocyte, 
nuclei  begin  to  enlarge;  two  nuclei  in  the  edge  of  an  oocyte  can  be  seen  in  the  upper  portion  ol 
the  figure.  The  increased  size  is  evident.  Bar  represents  20  nm. 

FIGURE  3.  Nucleus  from  a  living  preparation  of  a  mnllinncleale  oocyle  of  F.  />y</iii<iciis. 
Note  the  very  large  and  irregular  aggregates  of  nucleoli.  This  nucleus  had  lampbrtish  chromo- 
somes. In  the  same  oocyte  there  were  other  nuclei  of  smaller  size.  Bar  represents  20  /mi. 

FIGURE  4.  Higher  magnification  of  nucleolar  aggregates  from  a  nucleus  of  a  multinucleate 
oocyte  of  F.  pyi/inacus.  These  aggregates  are  comparable  to  those  of  Figure  o.  Bar  represents 
5  fj.m. 

FIGURE  5.  Lampbrush  chromosome  from  a  living  preparation  <»l  a  nucleus  in>m  a  multi- 
nucleate  oocyte  of  F.  pyyinacits.  Bar  represents  5  /un. 

FIGURE  6.  Cross-section  of  an  oocyte  of  F.  pyt/nnicns  with  nuclei  of  various  sizes.  Large 
nuclei  are  located  towards  the  periphery  and  small  nuclei  are  clustered  in  the  center.  Lamp- 
brush  chromosomes  are  found  in  the  larger  nuclei.  Nucleoli  can  be  seen  in  most  nuclei.  Bar 
represents  20  /urn. 

FIGURE  7.  Autoradiogram  showing  the  incorporation  of  ::H-nridine  into  the  nuclei  of  a 
multinucleate  oocyte  of  F.  pyc/iuaciis.  There  was  incorporation  in  both  large  and  small  nuclei 
and  in  the  nuclei  of  follicle  cells.  Bar  represents  50  ^.m. 


204 


E.    M.   UKI.   IMNU   AM)   A.   A.    1 1  I'M  I'll  KI  KS 


T.\m.h    I  I 


Average  number  of  nuclei-  in  the  ovarian  oocyte.      The  number  in  parentheses  indicates  the  number  of 
oocytes  analyzed.     In  oocytes  of  250-500  /u;»  diameter,  only  the  large  nuclei  were  counted  (see  text). 


Diameter  of  oocyte 
(nm) 

Nuclear  diameter 
(yum) 

Number  of  nuclei/oocvte 
(±s.d.) 

Volume  oocyte/ 
volume  nuclei 
(±s.d.) 

Flectonotus  ftyguiaeus  (nine  frogs) 


100-  190  (6) 

6.1 

2013  ±  961 

4.1  ±  1.8 

200-  290  (9) 

13.9 

1484  ±  541 

3.9  ±  1.8 

300-  390  (4) 

21.5 

877  ±  95 

4.4  ±0.7* 

460     (2) 

28.5 

818 

4.9 

950     (1) 

180 

3 

49.0 

1200-1300  (3) 

393-274** 

1 

86.3 

1  500     (5) 

390-289 

1 

114.2 

2000-3000  (3) 

511-186 

1 

1186.3 

Gastrotheca  ovifera  (one  frog) 


200-  250  (2) 

30 

194.5 

1.9 

300-  400  (2) 

32 

411.5 

2.9 

410-  500  (5) 

40 

481.6 

3.6 

510-  570  (3) 

40 

390.0 

6.7 

Gastrotheca  sp.  (one  frog) 


200-  300  (4) 

33 

71.0 

9.1 

300-  400  (2) 

40 

77.0 

10.3 

500-  550  (3) 

33 

125.7 

32.7 

680     (1) 

40 

107.0 

45.9 

*  The  volume  of  2000  small  nuclei  (650  X  103  ju3)  was  added  to  the  nuclear  volume  of  each  oocyte 
since  small  nuclei  were  not  counted. 

**  Largest  and  smallest  diameters  of  nuclei  of  oval  shape. 

Once  meiosis  begins,  there  are  no  more  mitotic  divisions  within  a  given  cyst. 
The  beginning  of  meiosis  is  characterized  by  moderate  enlargement  of  the  oocyte 
and  its  nuclei.  The  nuclear  size  is  generally  uniform  and  the  distribution  of  nuclei 
seems  to  be  random,  except  in  oocytes  with  few  nuclei;  in  the  latter  case  most 
nuclei  are  located  toward  the  periphery.  Inside  each  nucleus  the  chromosomes  are 
visible  as  fine  threads,  and  there  are  one  to  a  few  nucleoli  of  rounded  shape. 
Elimination  of  nuclei  may  begin  at  this  early  stage,  judging  by  the  presence  of 
numerous  small,  pycnotic  nuclei  that  stain  darkly  with  hematoxylin ;  these  are 
found  among  more  normal  appearing  nuclei  located  centrally  in  the  oocyte  (Figs. 
1,6). 

As  the  oocyte  grows,  there  develop  conspicuous  differences  in  nuclear  size, 
correlated  with  their  location  in  the  oocyte.  Nuclei  located  toward  the  periphery 
enlarge  to  a  greater  extent  than  those  located  more  centrally.  These  differences 
can  be  detected  in  oocytes  of  250  ,u,m  diameter  and  are  soon  obvious.  In  oocytes  of 
300  fj.m  the  largest  nuclei,  located  just  beneath  the  cortex,  measure  about  20 


MULTINUCLEATE  OOGENESIS  205 

nuclei  located  somewhat  deeper  are  .smaller;  those  located  in  the  center  are  the 
smallest  (6  to  7  p.m)  (Figs.  1,  6).  Large  and  medium  size  nuclei  contain  chromo- 
somes in  the  lampbrush  state  (Fig.  S)  ;  in  smaller  nuclei  the  chromosomes  are 
visihle  as  fine  threads  but  other  characteristics  could  not  be  determined.  The 
nucleoli  of  each  nucleus  seem  to  associate  into  a  few  large  masses  of  irregular  shape, 
nucleolar  aggregates  (Figs.  3,  4),  which  are  common  to  all  nuclei,  except  possibly 
those  that  are  pycnotic.  Growth  of  the  oocyte  is  accompanied  by  further  enlarge- 
ment of  the  peripheral  nuclei  (Table  II)  and  a  decrease  in  their  number.  Corre- 
lated with  the  smaller  number  of  large  nuclei  there  seems  to  be  an  increase  in  the 
number  of  medium  size  nuclei  and  small  nuclei.  The  latter  nuclei  are  very  abun- 
dant in  oocytes  of  300  to  700  /xin  diameter  and  are  clustered  in  the  center  of  the 
oocyte  (Fig.  6)  ;  they  are  so  numerous  and  closely  packed  that  they  are  impossible 
to  count,  but  their  number  seems  to  vary.  Owing  to  the  difficulty  of  counting  the 
centrally  located  nuclei,  only  the  larger  nuclei  were  counted  in  oocytes  of  250  to  500 
p.m  diameter  (Table  II).  The  decrease  in  total  number  of  nuclei  during  this  period 
is  thus  more  gradual  than  is  suggested  by  the  figures  in  Table  II,  which  shows  ex- 
clusively the  decrease  in  the  number  of  larger  nuclei.  Indeed,  the  number  of  nuclei 
appears  to  remain  rather  constant  from  the  formation  of  the  oocyte  up  to  a  size  of 
about  400  /mi  and  possibly  later.  In  one  oocyte  of  390  /mi  diameter,  for  example, 
the  number  of  small  nuclei  was  estimated  to  be  about  2500  with  a  mean  diameter  of 
8.6  /mi ;  in  the  same  oocyte  there  were  about  750  large  nuclei  with  diameters  of 
about  25  ju,m. 

In  oocytes  of  700  to  800  /mi  the  chromosomes  are  no  longer  in  the  lampbrush 
condition  and  cannot  be  seen  in  most  nuclei.    All  nuclei,  including  the  smallest  ones, 
contain  large  and  conspicuous  nucleolar  aggregates  of  irregular  shape.     In  oocytes 
of  800  to  900  /xm,  the  number  of  large  and  small  nuclei  decreases  markedly  (Figs.  8. 
9)  ;  during  this  process  the  large  nuclei  decrease  somewhat  in  volume  and  become 
pycnotic  before  finally  disappearing.     Associated  with  nuclear  disappearance  there 
are  changes  in  the  appearance  of  the  cytoplasm,  presumably  the  result  of  the  addi- 
tion of  nuclear  material   (Figs.   1,  8,  9).     This  modified  cytoplasm  is  originally 
distributed  unevenly  in  the  oocyte,  but  the  several  areas  of  differing  appearance 
eventually  join  and  give  a  more  homogeneous  character  to  the  cytoplasm.     Numer- 
ous vesicles  are  formed  slightly  later  and  the  oocyte  eventually  becomes  almost 
filled  with  them.    When  only  1  to  10  large  nuclei  remain  in  the  oocyte,  yolk  platelets 
begin  to  appear.    At  this  time  the  diameter  of  the  oocyte  is  approximately  1000  /mi. 
The  remaining  nuclei  enlarge  greatly  and  reach  a  diameter  of  about   180  /j.m. 
These  large  nuclei  occupy  the  central  region  of  the  oocyte,  in  contrast  to  the  periph- 
eral location  of  large  nuclei  during  earlier  stages.     After  oocytes  reach  about  2 
mm  in  diameter,  only  one  nucleus  is  found.     This  germinal  vesicle  is  originally 
spherical  and  is  located  near  the  center  of  the  cell  (Fig.  10)  ;  as  vitellogenesis  pro- 
ceeds, it  moves  toward  the  periphery  of  the  oocyte  and  becomes  ovoid   (Fig.   11  ). 
It  is  considerably  larger  than  any  of  the  preceding  nuclei. 

In  the  last  of  the  multiple  nuclei  and  in  the  germinal  vesicle  there  are  sometimes 
accumulations  of  basophilic  material  just  inside  the  nuclear  envelope  (Fig.  11)  ; 
the  nature  of  this  material  is  unknown.  In  each  of  the  final  nuclei,  the  nucleolar 
aggregates  become  very  large  before  disappearing  (Fig.  10)  ;  these  seem  to  be  re- 


206 


]•:.   \\.  DEL  PINO  AND  A.  A.  HUMPHRIES 


..  * 

-. 

*!*'  »         fcJ1.     -•':*  .* 


FIGURE  8.  Cross  section  of  ovary  of  F.  pyyutacus.  In  the  two  upper  oocytes  there  is  great 
reduction  in  the  number  of  nuclei.  Large  nuclei  shrink  and  become  pycnotic  before  disappearing. 
Both  oocytes  show  also  the  uneven  distribution  of  cytoplasm  modified  by  nuclei.  In  the  lower 
portion  of  the  figure  there  is  a  multinucleate  oocyte  with  nuck-i  of  various  sizes  and  a  follicle  of 
atresia.  Bar  represents  100  /*m. 


MULTINUCLEATE  OOGENESIS  207 

placed  gradually  l>v  small  ovoid  corpu>cles  that  might  In-  nucleoli.  Jii  addition, 
there  are  many  spherical  corpuscles,  much  smaller  than  the  usual  nucleoli  (Fig. 
11).  Both  types  of  corpuscles  are  present  in  the  germinal  vesicle  of  the  largest 
oocytes.  The  chromosomes  of  this  stage  were  not  found. 

The  yolk-filled  cytoplasm  of  large  ovarian  oocytes  and  ovulated  eggs  contains 
numerous  transparent  round  vesicles  that  are  distributed  almost  uniformly.  The 
animal  pole  of  the  egg  is  almost  devoid  of  both  vesicles  and  yolk  granules;  it  is 
within  this  zone  that  the  germinal  vesicle  is  located.  After  ovulation,  the  germinal 
vesicle  could  not  be  found  in  the  onlv  such  egg  that  was  available,  suggesting  that 
ovulation  and  germinal  vesicle  breakdown  occur  at  about  the  same  time. 

Small  nuclei  of  large  oocytes  gave  a  strong  Feulgen  reaction,  while  lampbrush 
chromosomes,  nucleoli  and  nucleolar  aggregates  of  larger  nuclei  reacted  weakly. 
Although  nucleolar  aggregates  did  not  ordinarily  give  a  po>itive  reaction,  there  was 
sometimes  a  positive  reaction  from  large  bodies  that  seemed  to  be  nucleolar  aggre- 
gates. In  the  germinal  vesicle,  only  the  ovoid  corpuscles  reacted  with  Feulgen,  and 
these  only  faintly. 

All  nuclei  in  oocytes  up  to  about  500  //in  in  diameter  incorporated  3H-uridine 
(Fig.  7).  Uridine  incorporation  was  detected  in  association  with  chromosomes 
during  the  early  stages  of  meiosis  ;  later,  however,  undine  seemed  to  be  incorporated 
almost  exclusively  by  the  nucleoli.  In  large  oocytes.  where  the  nuclear  number  was 
greatly  decreased,  there  was  slight  or  no  incorporation  of  label. 

Xuclcar  changes  during  oogenesis  in  Gastrotheca  ovifera  and  otJicr  marsupial  frogs 

During  oogenesis  the  oocyte  becomes  multinucleate  in  G.  oi'ifcra  and  other 
species  of  marsupial  frogs  (Table  III  ),  but  the  number  of  nuclei  per  oocyte  in  these 
frogs  is  smaller  than  that  in  F.  p\<jmacus  (Table  ID.  In  G.  orifera  the  nuclear 
number  reaches  about  500,  presumably  the  result  of  approximately  nine  mitotic 
divisions  of  the  original  oogonium  ;  in  Gastrotheca  sp.  (Venezuela)  there  are  only 
about  100  nuclei,  corresponding  to  fewer  than  seven  divisions.  The  number  of 
nuclei  in  oocytes  of  other  species  with  multinucleate  oocytes  (Table  III)  was  not 
counted,  but  it  seems  comparable  to  that  of  G.  oi'ifcra  or  Gastrotheca  sp.,  with  the 
exception  of  Pfciniphractus  jolnisoni,  which  has  fewer  nuclei.  There  was  a  maxi- 
mum of  four  nuclei  per  oocyte  in  the  only  ovary  of  this  species  that  was  examined. 

FIGURE  9.  Detail  of  the  changes  in  large  nuclei  prior  to  disappearance.  The  cytoplasm  ad- 
jacent to  nuclei  becomes  modified  and  the  nuclei  shrink  before  becoming  pycnotic.  Bar  repre- 
sents 50  fim. 

FIGURE  10.  Cross  section  of  the  nucleus  from  a  mononucleate  oocyte  of  F.  pyr/macus. 
This  nucleus  has  enlarged  greatly  and  shows  very  large  aggregates  that  will  be  replaced  later 
by  smaller  corpuscles.  Bar  represents  25  p.m. 

FIGURE  11.  Cross  section  of  the  single  germinal  vesicle  of  an  oocyte  of  F.  pyninacus. 
There  are  ovoid  corpuscles  towards  the  center  and  very  abundant  spherical  entities  of  smaller 
size.  The  chromosomes  could  not  he  seen.  Note  the  basophilic  accumulations  of  material 
just  inside  the  nuclear  envelope.  Bar  represents  25  nm. 

FIGURE  12.  Cross  section  of  a  cyst  in  the  ovary  of  Gastrotheca  oi'ifcra.  Most  oogonia  are 
mononucleate ;  these  will  fuse  to  give  a  multinucleate  oocyte  with  about  500  nuclei.  Bar  repre- 
sents 50  ,um. 

FIGURE  13.  Cross  section  of  an  oocyte  of  G.  mijern.  The  enlarged  nuclei  are  distributed  in 
a  layer.  Nuclei  are  absent  from  the  central  region.  Bar  represents  100  /j.m. 


208 


E.  M.  DEL  PINO  AND  A.  A.  HUMPHRIES 


TABLE  1 1 1 

Type  of  oogenesis  and  development  at  birth  in  marsupial  frogs. 


Species 

Egg  diameter 
(mm) 

Development 

Tadpole 

Froglet 

Mononucleate  type  of  oogenesis 


Cryptobatrachus 

C.  fi'.hrniinnii 
Flectonotus 

F.  fis  silis 
Gastrotheca 

G.  argentcoi'ircns 
G.  cai'ia 

G.  :hristiani 
G.  excnbitor 
G.  grttcilis 
G.  lojaiia 
G.  iiiarsupiata, 
G.  nicrtensi 
G.  monticola 
G.  ochoai 
G.  pcruana 
G.  phtinbea 
G.  riobambae 
G.  tcstndinca 
G.  sp.  (LSUMZ-32049) 
G.  sp.  (FMNH-39889) 
Hemiphractus 
//.  biibaliis 
H.  fasciatns 
H.  proboscideus 
H. 


3 
4 
6 
3 

2.5 
5 

5 
3 
4 
3 

3.5 


7 
10  (2) 


+  (1) 


+  (1) 


Alultimicleate  type  of  oogenesis 


Amphignathodon 

A.  giienthcrl 

+ 

Flectonotus 

F.  pygmaens 

3 

+  (3) 

Gastrotheca 

G.  ceratophrys 

8 

+ 

G.  cornuta 

+ 

G.  grisivoldi 

8 

+ 

G.  microdisca 

+ 

G.  ovifera 

8 

+ 

G.  weinlandii 

+ 

G.  sp.  (Venezuela) 

7 

+ 

Hemiphractus 

H.  johnsoni 

+ 

Stefania 

S.  sea  la  e 

9 

+ 

Total 

12 

21 

Number  of  species:  33.      (1)  Young  are  born  as  froglets  or  towards  the  end  of  metamorphosis. 
(2)  Egg  diameter  from  Trueb,  1974.      (3)  Born  as  advanced  tadpole. 


MULTINUCLEATE  OOGENESIS  2()(> 

In  (/.  ovijcru,  oogonial  divisions  within  a  cvst  are  apparently  synchronous,  as  in 
F.  p\'(/inaens,  and  most  oogonia  contain  a  single  nucleus  (Fig.  12).  Oogonia  with 
2  to  4  nuclei  are  rare,  in  contrast  to  F.  pyyniacns,  where  many  oogonia  seem  to  be 
multinucleate.  In  G.  oi'ifcra  the  oocyte  seems  to  lie  formed  by  fusion  of  mono- 
nucleate  oogonia  of  a  cyst.  \Yith  the  onset  of  meiosis,  nuclei  enlarge,  form  lamp- 
brush  chromosomes  and  later,  conspicuous  nucleolar  aggregates.  Most  nuclei, 
and  possibly  all  of  them,  become  arranged  toward  the  peripheral  region  of  the 
oocyte,  forming  one  or  two  layers  of  nuclei  (Fig.  13).  The  nuclei  of  the  internal 
layer  are  slightly  smaller  than  those  toward  the  periphery.  In  addition,  there  are 
a  few  pycnotic  nuclei  of  small  size  that  are  found  just  internal  to  the  larger  nuclei. 
Pycnotic  nuclei  have  also  been  observed  in  the  periphery  and  sometimes  in  the 
central  region  of  the  oocyte.  In  most  cases,  however,  there  are  no  nuclei  in  the 
central  part. 

As  the  oocyte  grows,  there  is  a  gradual  disappearance  of  nuclei  and,  toward  the 
beginning  of  vitellogenesis,  there  is  apparently  a  period  of  more  rapid  nuclear  dis- 
appearance. The  onset  of  vitellogenesis  is  accompanied  by  changes  in  the  cytoplasm 
of  the  oocyte  that  are  comparable  to  the  changes  described  for  F.  pygmacns. 
Gastrotheca  oz'ifcra  produces  very  large  oocytes  (approximately  8  mm  diameter) 
and  it  seems  that  the  oocyte  contains  multiple  nuclei  until  a  large  size  is  attained : 
oocytes  of  about  2  mm  in  diameter  do  not  have  yolk  platelets  and  still  have  numer- 
ous large  nuclei. 

Eleven  of  the  33  species  of  marsupial  frogs  examined  (Table  III)  are  char- 
acterized by  the  presence  of  multiple  nuclei  in  the  oocyte  during  the  early  stages  of 
oogenesis.  Although  the  formation  of  the  oocyte  in  some  of  these  species  was  not 
observed,  those  that  were  studied  show  similarities  to  F.  pyginacus  or  G.  ovijcra. 
In  the  majority,  the  nuclei  become  arranged  toward  the  peripheral  region  of  the 
oocyte  and  enlarge.  The  central  region  does  not  have  nuclei  in  most  instances,  but 
occasionally  a  few  small  pycnotic  nuclei  can  be  seen.  Each  nucleus  contains  lamp- 
brush  chromosomes  and  nucleolar  aggregates.  The  number  of  nuclei  decreases 
gradually  as  the  oocyte  grows ;  the  multinucleate  condition,  however,  is  prolonged 
into  the  vitellogenic  period,  as  in  G.  oi'ifcra. 

DISCUSSION 

There  is  little  doubt  that  the  multiple  entities  described  here  as  nuclear  are 
either  nuclei  or  derivatives  of  nuclei.  The  presence  of  visible  chromosomes  in 
many  of  the  nuclei  and  the  incorporation  of  labeled  uridine  attest  to  this  view,  as  do 
ultrastructural  studies  (in  preparation)  which  reveal  the  presence  of  a  typical 
nuclear  envelope.  What  is  unclear  is  whether  the  nuclei  all  contain  chromosomes 
and  whether  the  number  of  chromosomes  in  each  is  the  same.  The  extreme  varia- 
tion in  size  of  the  nuclei  suggests  that  the  chromosome  number  may  not  be  the  same, 
or  that  the  condition  of  the  chromosomes  mav  varv  from  the  large  nuclei  to  the 

»  * 

small.  Although  no  information  is  available  as  to  the  DNA  content  of  these  multi- 
nucleate  oocytes  or  the  DNA  content  of  individual  nuclei,  the  situation  suggests  a 
massive  overall  amplifiation  of  the  oocytes'  DNA,  considerably  greater  than  that 
occurring  in  the  eight-nucleate  oocytes  found  in  Asca pints  Intel  (Macgregor  and 
Kezer,  1970).  All  nuclei  are  presumably  engaged  in  synthesizing  RNA,  at  least 
prior  to  the  final  stages  where  only  a  single  germinal  vesicle  or  a  few  large  nuclei 
are  present.  It  appears,  then,  that  these  multinucleate  oocytes  represent  a  remark- 


210  E.  M.  DEL  PINO  AND  A.  A.  HUMPHRIES 

able  case  of  a  sort  of  endopolyploidy  of  high  degree,  even  if  the  nuclei  contain  ab- 
normal chromosome  complements. 

We  have  no  explanation  to  propose  for  the  widespread  occurrence  of  multi- 
nucleate  oocytes  in  marsupial  frogs ;  moreover,  we  have  been  unable  to  find  a  clear 
relationship  between  the  occurrence  of  the  multinucleate  condition  in  these  species 
and  such  characteristics  as  the  final  size  of  the  egg  or  the  pattern  of  reproduction. 
However,  there  does  seem  to  be  some  correlation  between  the  multinucleate  condi- 
tion, production  of  unusually  large  eggs,  and  development  of  young  to  the  froglet 
stage  in  the  maternal  pouch.  A  notable  exception  is  F.  pyymacus,  in  which  the 
eggs  do  not  seem  unusually  large  and  in  which  the  young  are  kept  in  the  pouch 
only  until  they  are  advanced  tadpoles.  An  egg  of  3  mm,  however,  might  be  con- 
sidered large  for  such  a  small  frog  (30  mm  snout-vent  length),  since  other  frogs 
with  multinucleate  oogenesis  measure  from  40  to  80  mm  (snout-vent  length).  Some 
frogs  with  mononucleate  oogenesis,  however,  are  of  similar  size  and  produce  eggs 
that  are  as  large  as  those  of  species  with  multinucleate  oocytes.  In  both  the  multi- 
nucleate  and  mononucleate  groups  there  are  species  that  give  birth  to  tadpoles  and 
species  that  give  birth  to  froglets  (Table  III).  One  generalization  that  may  be 
important  is  that  the  marsupial  frogs,  regardless  of  nuclear  number  during  oogene- 
sis, produce  eggs  that  are  very  large  by  usual  amphibian  standards,  but  very  few  in 
number.  Furthermore,  reproduction  seems  to  be  exceptionally  efficient :  almost  all 
eggs  produced  appear  to  be  fertilized  and  undergo  development.  In  F.  pyginacus, 
for  example,  the  number  of  large  eggs  in  the  ovary  is  small,  and  corresponds  closely 
to  the  number  of  embryos  in  the  pouch. 

The  total  nuclear  volume  increases  considerably  during  early  oogenesis,  but  the 
concomitant  great  increase  in  oocyte  volume  results  in  what  seems  to  be  a  rather 
constant  ratio  between  the  two  during  the  previtellogenic  period  (Table  II).  No 
firm  conclusions  can  be  reached,  however,  since  the  estimations  of  volume  are  sub- 
ject to  large  error  and  since  there  is  also  great  variability  in  the  number  of  nuclei 
per  oocyte.  During  the  vitellogenic  phase,  oocyte  growth  is  not  paralleled  by  an 
increase  in  nuclear  volume ;  thus,  the  ratio  of  oocyte  to  nuclear  volume  increases 
greatly  toward  the  end  of  oogenesis  (Table  II).  Little  information  is  available 
regarding  ratios  between  oocyte  volume  and  nuclear  volume  in  the  Amphibia,  but 
in  the  mononucleate  oocytes  of  Rana  pipiens  (Parmenter  ct  al.,  1960)  the  ratio 
seems  to  increase  earlier  than  in  F.  p\ymacus.  In  any  case,  however,  the  multi- 
nucleate  condition  obviously  results  in  an  enormous  nuclear  surface  area,  particu- 
larly in  oocytes  of  F.  pyymacus,  with  their  large  number  of  centrally  located  nuclear 
entities. 

The  multinucleate  condition  in  the  oocytes  described  in  this  report  differs  in 
several  notable  ways  from  that  in  Ascaphus  truei  (Macgregor  and  Kezer,  1970). 
The  maximum  number  of  nuclei  in  Ascaphiis  oocytes  is  usually  eight,  although 
many  ovaries  contain  a  few  oocytes  with  sixteen  nuclei  (Kezer,  personal  com- 
munication) ;  the  condition  arises  from  failure  of  cytokinesis  to  occur  following 
nuclear  division.  In  F.  pygmaciis,  G.  orijcra,  and  Gastrotheca  sp.,  however,  the 
number  of  nuclei  reaches  the  hundreds,  and  the  condition  seems  to  originate  through 
the  disappearance  of  cell  membranes  between  adjacent  cells  within  a  cyst.  Another 
difference  is  the  fact  that  in  Ascaphiis  the  multinucleate  condition  persists  until  late 


MULTINUCLEATE  OOGENESIS  211 

oogenesis,  while  in  at  least  some  of  these  marsupial  frogs  the  multinucleate  condi- 
tion lasts  only  until  ahout  the  time  yolk  formation  begins.  In  both  situations,  how- 
ever, the  multiple  nuclei  contain  chromosomes  in  the  lampbrush  condition,  and  there 
are  similar  nucleolar  features.  In  neither  situation  is  there  good  evidence  as  to  the 
mechanism  bv  which  the  number  of  nuclei  is  finallv  reduced  to  one. 


This  work  would  have  been  impossible  without  help  in  the  provision  of  living 
and  preserved  specimens  of  these  frogs  and  without  the  availability  of  museum  ma- 
terials for  analysis.  For  the  collaboration  we  received,  we  acknowledge  with 
gratitude  the  help  of  the  following  persons :  Dr.  William  E.  Duellman,  who  pro- 
vided specimens,  arranged  loans  from  various  museums  and  arranged  a  visit  to  the 
Museum  of  Natural  History  of  the  University  of  Kansas  for  E.  M.  del  Pino.  His 
recent  herpetological  work  in  South  America  was  supported  by  grants  NO.  GB- 
42481  and  DEB76-09986  from  the  National  Science  Foundation.  Dr.  Raymond  F. 
Laurent  from  Fundacion  Miguel  Lillo,  Tucuman,  Republica  Argentina,  provided 
working  space  at  that  institution  for  E.  M.  del  Pino  and  allowed  the  study  of 
museum  specimens  of  Gastrothcca.  Mr.  Scott  J.  Maness  from  Estacion  Biologica 
de  Rancho  Grande,  Republica  de  Venezuela,  provided  living  and  preserved  speci- 
mens from  that  area.  Dr.  Gonzalo  Medina  Padilla,  Jefe  de  la  Oficina  de  Fauna  del 
Ministerio  de  Agricultura  y  Cria,  gave  the  corresponding  permits  for  the  collection 
and  export  of  frogs  from  Venezuela.  Dr.  Richard  P.  Seifert  and  Florence  Hammet 
Seifert  aided  in  the  collection  of  frogs  from  Venezuela. 

This  work  was  supported  in  part  by  a  grant  from  the  McCandless  Fund  of 
Emory  University  and  by  grants  from  the  Regional  Program  PNUD/UNESCO  RLA 
75/047  and  RLA  76/006. 

SUMMARY 

The  occurrence  of  multinucleate  stages  during  oogenesis  appears  to  be  wide- 
spread in  the  marsupial  frogs  of  South  America.  In  some  species  the  number  of 
nuclei  or  nucleus-like  entities  per  oocyte  is  estimated  to  be  as  high  as  2000,  but  the 
number  in  other  species  may  be  considerably  lower ;  some  species  have  only  a 
single  oocyte  nucleus.  In  all  cases  it  seems  that  only  a  single  nucleus  remains  as 
the  oocyte  approaches  maturity.  The  situation  suggests  a  massive  general  amplifica- 
tion of  the  genome  of  the  multinucleate  oocytes  that  is  much  greater  than  has  yet 
been  reported.  Possible  relationships  between  the  occurrence  of  the  multinucleate 
condition  and  such  features  as  egg  size  and  reproductive  pattern  are  discussed,  but 
no  final  conclusions  can  be  made  on  the  basis  of  the  evidence  presently  available. 

LITERATURE  CITED 

DEL  PINO,  E.  M.,  AND  G.  SANCHEZ,  1977.  Ovarian  structure  of  the  marsupial  frog  Gastrothcca 
riohambac  (Fowler).  J.Morplwl..  153:  153-162. 

DEL  PINO,  E.  M.,  M.  L.  GALARZA,  C.  M.  DE  ALBUJA,  AND  A.  A.  HUMPHRIES,  JR.,  1975.  The 
maternal  pouch  and  development  in  the  marsupial  frog  Gastrothcca  riobainbae  (Fow- 
ler). Biol  Bull.,  149:  480-491. 


212  E.  M.  DEL  PINO  AND  A.  A.  HUMPHRIES 

DUEIXMAX,  \Y.  E.,  1976.    Liste  der  rezcnten  Amphibicn  und  Reptilien.     Hylidac,  Centrolenidae, 

Pseudidae.     Pages  1-225'  in  R.  Mertens,  W.  Hennig  and  H.  Wernuth,  Eds.,  Das  Tier- 

rcicli,  Licfcrung  95.    Walter  de  Gruyter,  Berlin. 

HUMPHREY,   R.   R.,   1963.     Polyploidy  in  the  Mexican  axolotl    (Ambystoma   mexicanum)    re- 
sulting from  multinucleate  ova.    Proc.  Nat.  Acad.  Sci.  U.S.A.,  50:  1122-1127. 
HUMPHRIES,  A.  A.,  JR.,  1956.    A  study  of  meiosis  in  coelomic  and  oviducal  oocytes  of  Triturus 

riridcsccns,  with  particular  emphasis  on  the  origin  of  spontaneous  polyploidy  and  the 

effects  of  heat  shock  on  the  first  meiotic  division.     /.  Morphol.,  99:  97-136. 
HUMPHRIES,   A.   A.,  JR.,   1966.     Exceptional  meiotic  conditions  in  oocytes  of  Ambystoma   /<//- 

poidcniii,  Triturus  viridcsccns  and  Xcnopus  lacvis  and  their  relation  to  the  origin  of 

spontaneous  heteroploidy.    Cytogcnetics,  5:  401-410. 

KING,  H.  D.,  1908.    The  oogenesis  of  Bitfo  !cnti</inosits.  J.  Morplwl.,  19:  369-438. 
MACGREGOR,  H.  C.,  AND  J.  KEZER,  1970.     Gene  amplification  in  oocytes  with  8  germinal  vesicles 

from  the  tailed  frog  Ascaphus  truci  Stejneger.    Chromosoma,  29:   189-206. 
MERTENS,   R.,   1957.     Zur   Naturgeschichte  des   venezolanischen    Reinsen-Beutel   frosches,   Gas- 

trothcca  ovifcra.  Zool.  Cart.   (Ncue  Folgc),  23:   110-133. 
PARMENTER,   C.  L.,   M.   DEREZIN,   AND   H.   S.   PARMENTER,    1960.     Binucleate   and   trinucleate 

oocytes  in  post-ovulation  ovaries  of  Rana  pipiais.    Biol.  Bull.,  119:  224-230. 
ROBINSON,  E.  S.,  E.  M.  STEPHENSON  AND  N.  G.  STEPHENSON,  1973.     Nuclear  constitution  of 

primary  oocytes  of  the  frog  Lciopelma  hochstcttcri  (Ascaphidae).     Copcia,  1973:   173- 

176. 
TRUEB,   L.,    1974.     Systematic   relationships   of  neotropical   horned   frogs,   genus   Hemiphractus 

(Anura,  Hylidae).     Occas.   Pap.   Mus.  Nat.  Hist.  Univ.  Kansas,  29:   1-60. 


Reference:  Biol.  Bull.,  154:  213-225.     (April,  1978) 


EVIDENCE  FOR  A  NONINTESTINAL  NUTRITIONAL  MECHANISM 
IN  THE  RHYNCHOCOELAN,  LIN  BUS  RUBER  * 

FRANK  M.  FISHER,  JR.*  AND  JOHN  A.  OAKS  3 
Marine  Biological  Laboratory,  Woods  Hole,  Massachusetts  02543 

The  majority  of  the  members  of  the  phylum  Rhynchocoela  are  found  free-living 
in  the  benthos  of  the  intertidal  and  subtidal  zones.  These  worm-like  organisms  are 
described  as  acoelomate  Bilateria  with  a  complete  digestive  system  (Hyman,  1951 ). 
Coe  (1943)  described  the  rhynchocoelans  (nemerteans)  as  carnivorous  in  their 
feeding  behavior,  and  in  some  littoral  communities  these  organisms  are  apparently 
the  most  abundant  predators  (Roe,  1970). 

The  function  and  morphology  of  the  digestive  system  of  a  number  of  rhyn- 
chocoels  has  been  extensively  examined  with  regard  to  participate  food  digestion  by 
Gontcharoff  (1948),  Jennings  (1960,  1962),  Jennings  and  Gibson  (1969)  and 
Gibson  (1970)  utilizing  light  microscopy  and  histochemical  techniques. 

The  ability  of  numerous  free-living  marine  invertebrates  to  remove  organic 
solutes,  present  in  relatively  low  concentrations,  from  the  environment  has  been 
extensively  investigated  by  Stephens  and  his  associates  (e.g.,  Stephens,  1964;  Reish 
and  Stephens,  1969).  Similar  absorption  phenomena  for  small  molecular  weight 
organic  nutrients  have  been  identified  in  the  rhynchocoelan,  Linens  rnber,  by 
Fisher  and  Cramer  (1967).  This  paper  extends  those  preliminary  reports  and 
indicates  that  the  epidermal  free  surface  is  the  major  site  for  absorption  of  these 
nutrients  from  their  environment. 

MATERIALS  AND  METHODS 
Source  and  maintenance  of  animals 

Specimens  of  Linens  ruber  (5  to  S  cm)  were  collected  at  low  tide  in  the  vicinity 
of  Manomet  Point,  Massachusetts,  and  maintained  in  the  laboratory  in  running  sea 
water  at  20—24°  C.  Animals  were  starved  while  being  maintained  in  artificial  sea 
water  lacking  organic  solutes  (Cavanaugh,  1964),  and  all  animals  shipped  to  Rice 
University  for  study  were  held  in  Instant  Ocean  (34%0;  Aquarium  Systems,  Inc.). 
All  experimental  animals  were  provided  with  a  coarse  sand  substrate  of  sufficient 
depth  to  cover  the  bottom  of  the  holding  containers. 

Solute  accumulation  techniques 

Rhynchocoelans,  being  vermiform  animals,  were  manipulated,  as  described 
below,  after  the  methods  of  Fisher  and  Read  (1971  )  ;  however,  artificial  sea  water 

1  Supported  in   part  by  grants   from   The   Moody   Foundation    (70-115)    and  the   National 
Science  Foundation  (GB-3447  and  PCM  77-09112). 

2  Present  address :    Biology  Department,  Rice  University,  Houston,  Texas  77001 

3  Present   address :     Department   of   Anatomy,    College   of    Medicine,    University   of   Iowa, 
Iowa  City,  Iowa  52242 

213 


214  F.  M.  FISHER  AND  J.  A.  OAKS 

was  substituted  for  the  balanced  saline  used  by  those  authors  for  elasmobraneh 
parasites.  On  removal  from  preincubation  in  artificial  sea  water  at  20°  C  for  30 
minutes  to  1  hour,  individual  samples  of  at  least  five  organisms  were  carefully 
blotted  on  hard  surfaced  filter  paper  and  transferred  to  14C-galactose,  14C-glucose, 
"C-leucine  or  14C-alanine  containing  incubation  mixture.  Unless  otherwise  stated, 
the  incubation  time  and  temperature  were  two  minutes  and  20°  C.  Following  the 
incubation  period  the  groups  of  worms  were  removed  from  the  radioactive  mixture, 
quickly  rinsed  three  times  in  large  volumes  of  artificial  sea  water,  blotted,  weighed 
on  a  torsion  balance,  and  placed  in  2.0  ml  of  70%  ethanol.  Worm  carcasses  were 
extracted  in  the  ethanol  with  intermittent  shaking  for  at  least  18  hours  at  room 
temperature  (22-26°  C).  Aliquots  of  the  extraction  ethanol  were  dried  on  ringed- 
copper  planchettes  and  counted  in  a  low  background  gas-flow  counter  for  10  minutes 
or  104  counts,  whichever  process  ended  first.  The  original  incubation  solution 
was  diluted  1  :  100  with  70%  ethanol  and  counted  in  a  similar  manner  to  determine 
the  specific  activity  of  the  test  medium.  Ethanol  extracted  worms  were  dried  to 
constant  weight  in  aluminum  foil  tares  at  100°  C  and  weighed  on  an  analytical 
balance.  Total  water  within  worms  of  a  single  group  was  estimated  from  wet 
weight  and  dry  weight  comparisons.  These  data  were  used  to  calculate  the  internal 
radiolabeled  substrate's  concentration  represented  by  ^moles/ml  worm  water. 

Chemical  assays 

Polysaccharide  was  determined  by  the  phenol :  HoSO^  method  of  Dubois,  Gilles, 
Hamilton,  Rebers  and  Smith  (1956)  and  the  modified  anthrone  method  of  Dimler, 
Schaffer  and  Wise  (1952).  Glucose  was  estimated  by  the  glucose  oxidase  method 
utilizing  the  Glucostat  Special  (Worthington  Biochemical  Corporation).  Protein 
was  isolated  by  trichloroacetic  acid  (TCA)  precipitation  of  homogenates  of  worm 
carcasses  and  estimated  by  the  colorimetric  method  of  Lowry,  Rosebrough,  Farr 
and  Randall  (1951). 

Chromotagraphic  analysis  of  worm  extracts  was  accomplished  using  the  follow- 
ing solvents  on  W7hatman  #1  paper:  first)  N-butanol :  ethanol :  acetone  :  water  = 
50:40:30:20  (descending)  (Gray  and  Frankel,  1954);  secondly,  N-butanol : 
propionic  acid  :  water  ==  63  :  31  :  44  (descending)  (Crowley,  1963):  and  thirdly, 
1-propanol :  ethyl  acetate  :  water  ==7:1:2  (ascending)  (Baar,  1954).  Carbohy- 
drates were  visualized  on  developed  chromatograms  with  the  alkaline  silver  nitrate 
reagent  of  Trevelyan,  Procter  and  Harrison  (1950).  Radioactive  areas  were 
localized  on  chromatograms  using  a  gas-flow  scanner.  These  areas  were  eluted 
from  chromatograms  by  a  technique  similar  to  that  of  Dimler  ct  al.  (1952),  and  the 
eluates  were  reduced  to  dryness  in  vacua  at  40°  C.  The  redissolved  residue  was 
co-chromatographed  with  authentic  standards  in  the  above  solvents  for  positive 
identification. 

Ultrastructural  techniques 

Specimens  of  Linens  ruhcr  were  placed  in  4°  C,  0.12  M  monobasic  sodium 
phosphate-sodium  hydroxide  buffer  at  pH  7.4  containing  6%  glutaraldehyde  plus 
3%  sucrose.  Subsequently  the  partially  fixed  worms  were  cut  into  cross  sections 
about  1  mm  thick  and  returned  to  the  fixative  for  three  to  six  hours.  Fixation  was 


NUTRITION    IN  LINEUS  RUBER 


215 


V. 

£ 


^ 

<b 


2.0 


\.o 


i 


Glucose 

Protein 

Polysaccharide 


0-8 


0.4 


20 


40 


60 


Minutes 


FIGURE  1.  The  accumulation  of  10"4  M  external  glucose  into  free  pool,  proteins  and  polysac- 
charide  of  L.  rubcr  with  time.  Glucose  in  the  free  pool  was  determined  by  glucose  oxidase, 
while  calculations  of  glucose  incorporated  into  protein  and  polysaccharide  were  based  upon 
specific  activity  of  radioglucose.  Data  is  based  on  the  mean  of  four  samples  per  time  interval. 

terminated  by  washing  in  buffer  containing  5%  sucrose.  After  washing  in  buffer 
for  12  to  18  hours,  the  tissue  was  post-fixed  for  45  minutes  in  1%  osmium  tetroxide 
plus  \%  sucrose  in  the  same  buffer.  Post-fixed  tissue  was  rinsed  in  tap  water, 
dehydrated  through  a  series  of  ethanol  and  propylene  oxide  solutions  and  embedded 
in  Epon  Epoxy  812  (Shell  Chemical  Co.).  Thin  sections  cut  on  diamond  knives 
were  mounted  on  bare  copper  grids,  stained  with  lead  and  uranium  salts  and  viewed 
in  the  Philips  300  electron  microscope. 

RESULTS 
Permeation  of  glucose 

Linens  ruber  incubated  in  sea  water  containing  10"4  M  glucose  rapidly  ac- 
cumulated that  hexose  into  the  free  pool  within  the  worm.  When  extracts  were 
examined,  the  concentration  of  glucose  inside  the  worm  determined  by  chemical 
analysis  was  ten  times  that  in  the  external  medium  after  an  incubation  period  of 
20  minutes,  and  within  60  minutes  the  internal  concentration  was  approximately 
17  times  the  original  concentration  in  the  surrounding  sea  water  (Fig.  1).  The 
initial  "free"  glucose  in  the  worms  was  less  than  2  X  10~4  mole  of  "worm  water." 
If  trace  amounts  of  14C-glucose  (0.1  /xCi//xmole)  were  added  to  the  incubation 
medium,  both  protein  and  polysaccharide  fractions  of  the  nemerteans  became 
labeled  and  the  amount  of  incorporation  of  the  carbon  from  hexose  into  these 
large  molecular  weight  compounds  increased  up  to  at  least  60  minutes  (Fig.  1). 

It  could  be  argued  that,  because  these  worms  possess  a  digestive  system,  glu- 
cose was  ingested  from  the  surrounding  sea  water  and  accumulation  was  occurring 
across  the  intestinal  epithelium.  If  these  worms  were  obtaining  labeled  glucose  by 


216 


F.  M.  FISHER  AND  J.  A.  OAKS 


swallowing  the  sea  water  medium,  one  would  expect   nonuniform  distribution  ot 
radiocarbon  along  tin-  linear  dimension  of  the  worm  body;  i.e.,  during  short  term 
incubation  periods  more  label  would  be  absorbed  into  tissues  near  the  mouth  and 
anterior  end  of  the  animal,  and,  as  the  length  of  the  incubation  period  increased, 
the  label  would  proceed  down  the  gut  toward  the  anus.     Individual  rhynchocoels 
were  incubated  for  varying  periods  of  time  in  sea  water  containing  10  4  M  glucose- 
14C.     At  the  termination  of  the  incubation  period  the  worms  were  rinsed  in  sea 
water,  placed  in  an  extended  position  on  a  thin  glass  plate  resting  on  solid  CO2, 
and  frozen  immediately.     Each  worm  was  cut  into  0.5  cm  pieces  and  each  section 
was  individually  extracted  in  ethanol  for  determination  of  the  radioactivity  along 
the  length  of  the  nemertean.     The  data  in  Table  I  indicate  that  there  was  uniform 
distribution  of  label  along  the  length  of  the  nemertean  following  different  incubation 
periods.    In  no  case  was  there  any  difference  in  the  amount  of  radiocarbon  in  either 
end  of  the  worm  which  suggests  that  over  the  time  period  examined,  movement  of 
sea  water  into  the  gut  via  the  mouth  and/or  the  anus  is  not  a  significant  factor  in 
determining  the  distribution  of  accumulated  glucose.     These  data  also  suggest  that 
glucose  accumulation  is  occurring  over  the  entire  surface  of  the  worm  and  that 
there  is  no  difference  in  the  rate  of  glucose  influx  along  the  body  of  the  rhyn- 
chocoelan.     Confirmation  that  the  worms  were  not  ingesting  the  incubation  medium 
was  obtained  by  ligaturing  ten  specimens  of  Linens  rubcr  posterior  to  the  mouth 
and  anterior  to  the  anus  with  4/0  silk  suture  prior  to  the  incubation  in  the  radio- 
active substrate.     No  difference  in  distribution  of  accumulated  radioglucose  could 
be  observed  between  ligatured  and  the  unligatured  control  animals    (Table   II). 
Amounts  of  radioglucose/body  section  is  similar  in  control  and  ligatured  animals 
(Table  II). 

The  initial  rate  of  glucose  accumulation  was  examined  during  two-minute  incu- 
bations using  glucose  concentrations  ranging  over  several  orders  of  magnitude. 
Glucose  influx  exhibits  initial  rate  saturation  kinetics  as  a  function  of  concentration 

TABLE  I 

Distribution  of  ethanol- soluble  counts /minute  along  the  body  of  Linens  ruber  after  incubation  for  dif- 
ferent periods.     Initial  external  glucose-14C  concentration  is  10  4  M ;  incubation  temperature,  20°  C. 


counts/minute 

1  min 

2  min 

4  min 

8  min 

16  min 

1  anterior 

90 

160 

201 

340 

873 

2 

73 

171 

260 

339 

734 

3 

89 

163 

220 

370 

759 

4 

69 

181 

229 

310 

819 

5 

91 

170 

248 

319 

840 

6 

83 

190 

271 

357 

839 

7 

94 

167 

209 

348 

790 

8 

71 

159 

219 

301 

763 

9 

75 

179 

210 

359 

793 

10 

84 

187 

254 

779 

11 

70 

220 

810 

12  posterior 

61 

791 

NUTRITION     IN    LlXEl'S  RUBER 


217 


TABLE  II 

Distribution  of  ethanol- soluble  counts  'minute  along,  the  body  of  ligatured  and  nnligtttnre'1  I. incus  ruber. 
Initial  external  glucosc-l*C  concentration  is  K>-*  M  ;  incubation  time.  5  minutes  at  20°  C. 


Number  of 
section 

x  counts/minute 
error  of  the 

(±  standard 
mean) 

Ligatured                   \Vitliout  ligature 

1  anterior 

263  ±  47 

258  ±  47 

2 

243  ±  30 

261  ±  51 

3 

199  ±  61 

241  ±  32 

4 

231  ±  27 

229  ±  26 

5 

246  ±  43 

263  ±  59 

6 

271  ±  59 

237  ±  23 

7 

251  ±  47 

231  ±  31 

8 

220  ±  17 

243  ±  35 

9 

237  ±  30 

240  ±  40 

10 

246  ±  38 

221  ±  26 

11 

251  ±  53 

279  ±  60 

12  posterior 

239  ±  37 
n  =  10 

264  ±  61 

n  =  15 

(Fig.  2).  Galactose  uptake  followed  a  similar  entry  pattern  (Fig.  2).  The  ac- 
cumulation of  glucose  is  inhibited  by  galactose  (66%  inhibition)  and  three  gly- 
cosides :  phlorizin  (93 c/c  inhibition) ,  quercetin  (71%  inhibition)  and  ouabain  (29c/( 
inhibition).  Phlorizin,  the  /3-glucoside  of  phloretin  and  /^-D-glucose,  is  a  potent 
inhibitor  of  glucose  permeation  in  a  number  of  systems  (see  Crane,  1960).  The 
permeation  of  glucose  was  not  inhibited  by  other  hexoses,  or  di-  and  trisaccharides 
including  levulose,  mannose,  X-acetylglucosamine,  cellobiose,  maltose,  trehalose, 
sucrose  and  melibiose.  Amino  acids,  fattv  acids  and  organic  acids  also  failed  to 


S  (MxICT7) 


10 


100 


FIGURE  2.  The  effect  of  external  substrate  concentration  on  the  permeation  of  glucose  and 
galactose  into  the  free  hexose  pool.  Data  is  based  on  the  mean  of  three  samples  per  concen- 
tration of  hexose. 


218 


F.  M.  FISHER  AND  J.  A.  OAKS 


(MxICT7) 


8 


FIGURE  3.  The  effect  of  external  substrate  concentration  on  the  accumulation  of  alanine 
and  leucine  into  free  pool  of  L.  rubcr.  Data  is  based  on  the  mean  of  three  determinations  per 
concentration  of  amino  acid. 

inhibit  glucose  permeation.  In  these  experiments  the  initial  external  glucose  con- 
centration was  10~4  M  ;  and  the  inhibitors,  1O3  M.  All  incubations  were  carried  out 
for  two  minutes  at  20°  C. 

Chromatographic  examination  of  ethanolic  extracts  of  worms  after  two-minute 
incubation  in  radioglucose  revealed  that  there  was  very  little  (<0.5%)  metabolism 
of  the  glucose  during  the  relatively  short  incubation  period.  If  nemerteans  are 
post-incubated  in  sea  water  without  glucose  following  a  two-minute  incubation  in 
radioglucose,  there  is  no  chromatographically  identifiable  glucose  "leakage"  into 
the  efflux  medium.  There  are,  however,  traces  of  succinate  which  appear  in  the 
surrounding  medium  during  this  second  incubation  period. 

Permeation  of  alanine  and  Icnc'mc 

The  uptake  of  alanine  and  leucine  by  Linens  rubcr  also  followed  saturation 
kinetics  (Fig.  3).  The  fate  of  the  accumulated  alanine  was  further  examined  using 
long-term  incubations  to  follow  the  possible  appearance  of  radiocarbon  into  protein. 
The  results  of  this  experiment,  expressed  as  /mioles/g  protein  is  seen  in  Figure  4. 
The  radiocarbon  skeleton  from  alanine  was  incorporated  into  protein  during  the 
60-minute  experimental  period.  Lotiger  incubation  times  were  not  examined. 

The  accumulation  of  alanine  and  leucine  \vere  not  influenced  by  50 :  1  ratios  of 
carbohydrates,  organic  acids  or  fatty  acids ;  however,  the  uptake  of  those  substrates 
was  inhibited  by  some  amino  acids  examined  (Table  III).  Diabasic  and  dicar- 
boxylic  amino  acids,  the  imino  acid  proline,  and  the  sulfonic  acid  derivative,  taurine, 


NUTRITION    IN  LI.VEUS  RUBER 


TABLE  III 

Inhibition  of  alanine  and  leucine  by  other  l-ainino  acids.     External  substrate  /  OIK  entration  is  5  X 
M  ;  inhibitor  concentration,  2  X  10~3  M;  incubation  time,  2  minutes  at  20°  C. 


Inhibitor 


Inhibition 
of  alanine  permeation 


Inhibition 
of  leucinc  permeation 

<',  ) 


Alanine 

97 

4') 

/3-Alanine 

11 

0 

Arginine 

0 

0 

Aspartic  acid 

0 

0 

Glutamic  acid 

0 

0 

Glycine 

64 

30 

Isoleucine 

10 

53 

Leucine 

53 

94 

Lysine 

0 

0 

Methionine 

47 

61 

Proline 

0 

0 

Tau  ri  ne 

0 

0 

Valine 

21 

4<> 

did  not  influence  the  uptake  of  either  alanine  or  leucine.  The  mutual  inhibition  of 
uptake  by  leucine  and  alanine  suggested  that  those  amino  acids  may  compete  for 
entry  through  the  same  membrane  site. 

Chromatographic  examination  of  extracts  from  two-minute  incubations  indicated 
that  there  were  no  detectable  metabolites  of  either  alanine  or  leucine  in  the  free  pool 
of  solutes  within  the  worm  bodies.  Similar  examination  of  the  incubation  medium 
revealed  that  there  were  no  metabolites  of  these  amino  acids  excreted  into  the  ex- 
ternal medium  during  the  two-minute  exposure  to  the  isotopes. 

MorpJwIoyy  of  (lie  evidential  and  epithelial  surfaces 

Two  interfaces  for  the  accumulation  of  organic  solutes  exist  on  most  free-living, 
aquatic  metazoans  :  the  epidermal  covering  and  the  epithelia  lining  the  digestive 


<= 


0.9 


0) 


0.3 


15 


60 


30  45 

Minutes 

FIGURE  4.     The  effect  of  time  on  the  incorporation  of  10  r>  M  external  alanine  into  protein 
uf  L.  rnbcr.    Data  based  on  the  mean  of  four  samples  per  time  interval. 


220 


IF.  M.  FISHER  AND  J.  A.  OAKS 


s 


X 


* 


i 


*fT   a t  !  4Bk  ^IP»"  _    3H! 

I;n;rKE  5.  Ultrastructure  of  the  external  surface  of  Linens  nihcr  epidermis.  Apical  cyto- 
plasm and  plasmalemma  of  the  free  surface  possess  cilia  which  extend  beyond  the  outer  limits 
of  the  brush  border  (BB),  defined  by  the  tips  of  the  microvilli  (arrows).  Note  the  close 
spacing  of  adjacent  microvilli.  Bar  equals  1  ,um. 

FIGURE  6.  Ultrastructure  of  the  Juminal  surfaces  of  Linens  nihcr  gut.  Apical  cytoplasm 
and  plasmalemma  of  the  intestinal  free  surface  possess  a  sparse  population  of  microvilli 
(arrows)  projecting  from  the  intestinal  surface.  Note  that  these  microvilli  are  not  arranged  in 
a  parallel  array  as  observed  at  the  epidermal  surface.  Bar  equals  1  /um. 


NUTRITION    IN  LINEUS  RUBER  221 

system.  Both  epithelial  surfaces  of  Linens  were  examined  for  structural  specializa- 
tion which  could  account  for  the  uptake  of  the  dissolved  compounds.  The  free 
surfaces  of  the  gut  epithelium  and  of  the  epidermis  possess  microvilli  and  cilia 
amplifing  their  free  surface  (Figs.  5,  6).  However,  the  microvilli  are  far  more 
numerous  and  regularly  distributed  on  the  apical  surfaces  of  epidermal  cells  than 
they  are  on  the  corresponding  surfaces  of  the  intestinal  epithelial  cells  (Fig.  5). 
The  close  register  of  microvilli  at  the  epidermal  surface  resembles  that  described 
for  the  brush  borders  present  at  the  surfaces  of  organs  and  tissues  modified  for 
absorptive  function.  In  contrast  to  the  epidermis,  the  distribution  of  microvilli  at 
the  surface  of  the  intestinal  epithelium  in  Linens  is  relatively  sparse  and  less  con- 
sistently organized  (Fig.  6)  resembling  the  surfaces  where  the  function  of  surface 
amplification  is  not  well  understood  (i.e.,  vertebrate  trachea,  Steinman,  1968;  or 
the  trematode  miracidium,  Wilson,  1969). 

DISCUSSION 

There  are  diverse  opinions  on  the  role  of  dissolved  reduced  carbon  compounds 
as  a  source  of  energy  for  marine  metazoans.  Stephens  (1967)  pointed  out  that 
the  concept  of  the  utilization  of  dissolved  organic  material  in  the  nutrition  of  aquatic 
animals  is  not  new  and  that  Putter  (1908a,  b)  first  advanced  the  idea  of  their  im- 
portance. Later  Krogh  (1931)  dismissed  this  notion  and  concluded  that  there  was 
no  substantial  evidence  to  support  Putter's  hypothesis. 

Investigators,  too  numerous  to  completely  list  here,  have  since  shown  that  vari- 
ous soft-bodied  marine  invertebrates  can  remove  dilute  organic  solutes  from  sea 
water  (i.e.,  Stephens,  1968),  some  at  concentrations  as  low  as  3  X  10"9  M  (Goreau, 
Goreau,  and  Yonge,  1971).  Johannes,  Coward  and  \Yebb  (1969)  criticized  the 
methodology  involved  in  most  uptake  studies  using  amino  acids  because  the  net 
efflux  of  free  amino  acids  has  seldom  been  measured.  The  simultaneous  movements 
of  such  compounds  into  and  out  of  biological  systems  is  well  documented.  "\Yilbrandt 
and  Rosenberg  (1961)  and  Johannes  ct  al.  (1969)  stress  that  there  is  a  net  loss 
of  free  amino  acids  during  most  uptake  experiments  involving  radioactive  sub- 
strates. However,  the  significant  increase  of  internal  free  glucose  in  Linens  rnbcr 
incubated  in  10"4  M  glucose  suggests  that  the  amount  of  glucose  lost  by  efflux  is 
relatively  small  in  comparison  to  the  accumulated  glucose  available  to  the  worm's 
metabolism.  Therefore,  some  organic  solutes  available  in  the  organism's  environ- 
ment could  serve  as  a  significant  source  of  nutrition  for  this  nemertean. 

During  the  two-minute  incubation  period  used  to  determine  initial  rates  of  glu- 
cose accumulation,  there  are  no  metabolites  of  glucose  excreted  into  the  incubation 
medium,  and  less  than  0.5%  of  the  radiocarbon  inside  the  worm  is  identifiable  as 
nonglucose  moieties.  During  a  longer  incubation,  however,  there  is  metabolism  of 
glucose  to  succinate.  This  latter  metabolite  subsequently  appears  in  the  external 
sea  water,  and  the  concentration  of  this  acidic  end  product  increases  with  extended 
incubation  periods.  Absorbed  glucose  is  readily  incorporated  into  the  nemertean's 
polysaccharide  and  almost  equivalent  amounts  of  radiocarbon  are  incorporated  into 
the  proteins  of  the  worm  body.  Preliminary  analysis  of  the  TCA  precipitable  frac- 
tion indicates  that  the  radiocarbon  is  present  primiarily  as  alanine  with  a  small 
amount  of  aspartic  acid  suggesting  that  one  pathway  of  glucose  metabolism  in 
L.  ruber  may  resemble  those  reported  for  a  number  of  parasitic  platyhelminths 


222  F.  M.  FISHER  AND  J.  A.  OAKS 

(Von  Brand,  196(>).  Tlir.sc  data  also  strongly  indicate  that  this  hexosc  absorbed 
from  the  surrounding  sea  water  is,  in  fact,  serving  as  an  energy  source  for  this 
organism.  It  should  he  emphasized  that  the  concentration  of  glucose  used  in  these 
experiments  is  consistent  with  the  values  reported  for  dissolved  carbohydrate  in 
oceanic  waters  (Wangersky,  1952;  Wangersky  and  ( iuillard,  1960;  \Yalsh,  1965a, 
b,  1966;  Walsh  and  Douglass,  1966). 

Absorbed  alanine  enters  the  metabolic  systems  of  the  worm,  since  it  is  readily 
incorporated  into  the  TCA  precipitable  fraction  during  incubation  periods  of 
moderate  length.  Although  preliminary  in  nature,  these  data  indicate  that  amino 
acids  from  the  surrounding  sea  water  can  and  do  serve  as  a  source  of  amino-nitro- 
gen  for  protein  synthesis.  The  concentration  of  substrates  used  in  these  experi- 
ments are  also  within  the  range  of  those  reported  for  oceanic  and  estuarine  waters 
(Adams  and  Richards,  1968;  Belser,  1959  and  1963;  Chau  and  Riley,  1966;  Siegel 
and  Degens,  1966;  Webb  and  Johannes,  1967). 

Stephens  and  Schinske  (1961)  described  the  removal  of  amino  acids  from  sea 
water  by  numerous  invertebrates  belonging  to  eleven  phyla.     In  their  experiments 
only  arthropods  failed  to  remove  such  solutes  from  the  surrounding  water.  The  pres- 
ence of  a  hard,  acellular,  relatively  impermeable  cuticle  on  the  exterior  of  arthro- 
pods and  an  epidermis  on  soft-bodies  marine  invertebrates  suggests  that  the  epi- 
dermis may  be  the  site  for  absorption  of  some  soluble  substances  from  sea  water. 
The  observed  morphology  of  the  external  surface  of  Linens  rnbcr  is  typical  of 
many  of  those   soft-bodied   organisms.      MacRae    (1967)    has   also   found   brush 
border  microvilli  amplifying  the  epidermal  surface  area  in  contact  with  sea  w7ater  on 
Turbellaria;    Lloyd    (1969)    and    Lane    (1963)    on    molluscs;    Little    and    Gupta 
(1968)  and  Norrevang  (1965)  on  pogonophorans ;  Potswald  (1971)  on  annelids; 
and   Menton  and   Eisen    (1970)    on   echinodenns.     Except  for   the   cilia   present 
among  the  brush  border's  microvilli  in  some  invertebrate  epidermises,  these  brush 
borders  resemble  those  on  tissue  surfaces  known  to  possess  high  rates  of  transport 
of  amino  acids  and  monosaccharides,  such  as  the  vertebrate  intestine,  the  proximal 
tubule  of  the  kidney  and  the  tegument  of  tapeworms  (reviewed  by  Lumsden,  1975  ). 
The  resemblance  of  these  surfaces  is  also  consistent  with  the  hypothesis  that  the 
epidermis  covering  Linens  rnbcr  may  be  the  site  of  absorption  of  nutrient  molecules 
present  in  sea  water.    The  epithelium  of  the  digestive  system  in  Linens  rubcr  also 
possesses  microvilli.    These  could  serve  as  a  second  site  of  solute  absorption  from 
sea  water,  even  though  they  are  fewer  in  number  on  the  intestinal  surface  than 
those  present  on  the  epidermis.     Jennings  (1969)  suggests  that  the  intestinal  epi- 
thelium serves  to  phagocytize  partially  digested  material  from  the  lumen  of  the  gut 
rather  than  as  a  primary  surface  for  nutrient  solute  transport.     In  line  with  that 
suggestion,  our  experiments,  involving  the  use  of  ligatured  and  unligatured  nemer- 
teans,  indicates  that  the  epidermis  investing  this  worm  is  the  most  important,  if  not 
the  sole  route,  in  solute  feeding. 

In  a  preliminary  report,  Fisher  and  Cramer,  (1967)  suggested  that  the  mem- 
brane transport  of  solutes  represented  a  new  feeding  mechanism  in  the  phylum 
Rhynchocoela.  We  have  shown  that  glucose  enters  L.  rnbcr  by  a  mediated  process. 
Our  data  and  that  of  Fisher  (unpublished)  suggest  that  the  accumulation  of  glucose 
is  competitively  inhibited  by  galactose.  Three  glucosides  also  inhibit  glucose 


NUTRITION    IN  LINEUS  RUBER  223 

permeation ;  however,  aniino  acids,  fatty  acids  and  organic  acids  are  without  effect. 
This  accumulation  process  for  glucose  can  be  described  as  an  active  transport  sys- 
tem (Fisher,  unpublished).  The  facts  that  the  concentration  of  chemically  deter- 
mined glucose  inside  the  worm  is  greater  than  that  in  the  external  sea  water,  that 
a  stereoisomer  of  glucose  inhibits  uptake  and  that  glucose  is  accumulated  against  a 
concentration  gradient  also  support  the  notion  that  this  is  a  mediated  process. 

Alanine  and  leucine  also  enter  L.  ntbcr  by  a  mediated  process  which  demon- 
strates saturation  kinetics.  Entry  of  these  compounds  is  inhibited  by  other  neutral 
amino  acids ;  however,  acidic  and  basic  amino  acids,  as  well  as  proline  and  taurine, 
do  not  inhibit  the  uptake  of  alanine  and  leucine.  Our  data  and  that  of  Fisher  (un- 
published) indicate  that  the  uptake  of  alanine  is  competitively  inhibited  by  leucine. 

The  undiminished  accumulation  of  glucose  in  ligatured  and  nonligatured  ani- 
mals, the  incorporation  of  glucose  into  polysaccharide,  the  synthesis  of  amino  acids 
from  hexose  with  subsequent  incorporation  into  protein,  the  incorporation  of  an 
absorbed  amino  acid  into  protein  fraction,  and  the  consistency  of  the  epidermal 
morphology  with  other  tissues  which  are  known  to  transport  solutes,  strongly  sup- 
port the  notion  that  this  surface  of  Linens  nibcr  serves  as  a  functional  feeding 
mechanism,  capable  of  metabolite  accumulation  from  sea  water  in  its  littoral 
habitat. 

SUMMARY 

1.  Linens  rubcr  rapidly  accumulates  glucose  from   sea  water  into  free  pools 
within  the  worm  concentrating  the  hexose  to  17  times  the  original  external  concen- 
tration (10~4  M)  in  one  hour. 

2.  Accumulated  glucose,  alanine  and  leucine  are  incorporated  into  protein,  and 
additional  glucose  is  incorporated  into  polysaccharide.     No  free  glucose,  alanine  or 
leucine  is  effluxed  during  two  minutes ;  however,  succinate,  derived  from  glucose, 
is  detectable  in  the  external  medium  when  the  incubation  time  is  extended. 

3.  The  demonstration  of  saturation  kinetics  for  both  glucose  and  galactose,  the 
partial  inhibition  of  glucose  entry  by  galactose  and  inhibition  of  glucose  accumula- 
tion by  phlorizin,  quercetin  and  ouabain  is  consistent  with  specific  sites  of  glucose 
transport. 

4.  Similar  kinetics  for  both  alanine  and  leucine  accumulation,  their  mutual  com- 
petition for  entry  and  the  inability  of  carbohydrates,  organic  acid,  and  fatty  acids  to 
influence  the  uptake  of  alanine  and  leucine  is  consistent  with  specific  transport  sites 
for  neutral  amino  acids. 

5.  Comparison   of  glucose   accumulation   by   whole   ligatured   and   unligatured 
worms,  as  well  as  along  the  length  of  unligatured  worms,  indicates  that  a  majority 
of  the  sites  of  entry  available  to  glucose  in  the  worm's  environment  is  through  its 
epidermis. 

6.  Ultrastructural  examination  of  free  epidermal  and  gut  luminal  surfaces  re- 
veal that  each  is  bounded  by  a  plasmalemma  with  a  surface  area  expanded  by  micro- 
villi.     The  surface  area  of  epidermis  is  greatly  increased  by  numerous  microvilli 
arranged  in  the  form  of  a  brush  border  and  is  greater  than  the  analogous  surface 
region  of  the  gut.    Presence  of  a  brush  border  is  characteristic  of  tissues  with  high 
rates  of  transport  function. 


224  F.  M.  FISHER  AND  J.  A.  OAKS 

LITERATURE  CITED 

ADAMS,  D.  D.,  AND  F.  A.  RICHARDS,  1968.     Dissolved  organic  matter  in  an  anoxic  fjord  with 

special  reference  to  the  presence  of  mercaptans.    Deep  Sea  Res.,  15:  471-481. 
BAAR,  S.,  1954.     Estimation  of  glucose  by  paper  partition  chromatography.     Biochent.  J.,  58: 

175-176. 
BELSER,  W.  L.,   1959.     Bioassay  of  organic  micronutrients   in   the  sea.    Proc.  Nat.  Acad.  Sci. 

US. A..  45:  1533-1542. 
BELSER,  W.  L.,   1963.     Bioassay  of  trace  substances.     Pages  220-231   in  E.  D.  Goldberg,  Ed., 

The  Sea,  ideas  and  observations  on  progress  in  the  study  of  the  seas,  I'ol,  2  Marine 

chemistry.   John  Wiley  and  Sons,  New  York. 
CAVANAUGH,  G.  M.,  1964.    Formulae  and  methods  of  the  Marine  Biological  Laboratory,  Vol.  V . 

Marine  Biological  Laboratory,  Woods  Hole,  87  pp. 
CHAU,  Y.  K.,  AND  J.   P.  RILEY,   1966.     The  determination  of  amino  acids  in  sea  water.    Deep 

Sea.  Res.,  13:  1115-1124. 
COE,  W.  R.,  1943.    Biology  of  the  nemertines  of  the  Atlantic  Coast  of  North  America.     Trans. 

Conn.  Acad.  Arts' Sci.,  34:  59-61. 

CRANE,  R.  K.,  1960.     Intestinal  absorption  of  sugars.     Physiol.  Rev.,  40:  789-825. 
CROWLEY,  G.  J.,  1963.     Studies  in  arthopod  serology.    I.  Changes  in  hemolymph  composition  as 

related  to  ecdysal  cycle.     Wasmann  J.  Bio!.,  21:  177-191. 
DIMLER,  R.  J.,  W.  C.  SCHAFFER,  AND  C.  S.  WISE,   1952.     Quantitative  paper  chromatography 

of  D-glucose  and  its  oligosacchardies.     Anal.  Client.,  24:   1411-1413. 
DUBOIS,  M.,  R.  R.  A.  GILLES,  J.  K.  HAMILTON,  P.  A.  REBERS,  AND  F.  SMITH,  1956.  Colori- 

metric  method  of  the  determination  of  sugars  and  related  substances.    Anal.  Chem.,  28: 

350-356. 
FISHER,  F.  M.,  JR.,  AND  N.  CRAMER,   1967.     New  observation  on  the  feeding  mechanisms  in 

Linens  ruber  (Rhynchocoela).     Biol.  Bull.,  133:  464. 
FISHER,   F.   M.,  JR.,  AND  C.   P.  READ,   1971.     Transport  of  sugars   in  the  tapeworm   Calliobo- 

tliriitin  vei-tieillatuni.   Biol.  Bull..  140:  46-62. 
GIBSON,  R.,  1970.     The  nutrition  of  Paranemertes  ferei/rina   (Rhynchocoela :    Hoplonermertea) . 

II.    Observations  on  the  structure  of  the  gut  and  proboscis,   site  and  sequence  of  di- 
gestion, and  food  reserves.    Biol.  Bull.,  139  :   92-106. 
GONTCHAROFF,  M.,   1948.     Note  sur  1'alimentation  de  quelques  nemertes.     Sci.   Prot/r.   Natur. 

Paris.  10 :  75-78. 
GOREAU,  T.  F.,  N.  I.  GOREAU,  C.  M.  YONGE,  1971.     Reef  corals:    autotrophs  or  heterotrophs. 

Biol.  Bull.,  141 :  247-260. 
GRAY,  H.  E.,  AND  G.  FRANKEL,  1954.  Fructomaltose,  a  recently  discovered  trisaccharide  isolated 

from  honeydew.    Science,  118:  304-305. 
HYMAN,  L.  H.,   1951.     The  invertebrates:    Platvhelininthes  and  Rvnchoeocls.    The  acoclomate 

Bilateria,  Vol.  II.    McGraw  Hill,  New  York,  460  pp. 
JENNINGS,  J.  B.,  1960.     Observations  on  the  nutrition  of  the  rhynochocoelan   Linens  ruhcr   (O. 

F.  Miiller).    Biol.  Bull.  122:  63-72. 
JENNINGS,  J.  B.,  1962.     A  histochemical  study  of  digestion  and  digestive  enzymes  in  the  rhyn- 

chocoelan.  Linens  rubcr  (O.  F.  Miiller).     Biol.  Bull.,  122:  63-72. 
JENNINGS,  J.  B.,  1969.    Ultrastructural  observations  on  the  phagocytic  uptake  of  food  materials 

by  the  ciliated  cells  of  the  rhynchocoelan  intestine.     Biol.  Bull.,  137:  476-485. 
JENNINGS,  J.   B.,  AND  R.  GIBSON,   1969.     Observations   on  the  nutrition  of  seven   species   of 

rhynchocoelan  worms.     Biol.  Bull.,  136:  405-433. 
JOHANNES,  R.  E.,  S.  J.  COWARD  AND  R.  L.  WEBB,  1969.     Are  dissolved  amino  acids  an  energy 

source  for  marine  invertebrates?    Cowp.  Bioehem.  Physiol.,  29:  275-281. 
KROGH,  A.,  1931.    Dissolved  substances  as  food  of  aquatic  animals.     Biol.  Rev.,  6:  412-442. 
LANE,  N.  J.,  1963.     Microvilli  on  the  external  surfaces  of  gastropod  tentacles  and  body  walls. 

Q.  J.  Microsc.  Set.,  104:  495-504. 
LITTLE,  C.,  AND  B.  L.  GUPTA,  1968.     Pogonophora :  uptake  of  dissolved  nutrients.  Nature,  218: 

873-874. 

LLOYD,  D.  C.,  1969.     Some  observations  on  the  skin  of  O.rycliilus  spp..    (Fitzinger),  with  par- 
ticular reference  to  O.  helvefieus   (Blum)    (Mollusca,   Pulmonata,  Aonitidae).  Proto- 

plasina,  68 :  327-339. 


NUTRITION    IN  LIXEUS  RUBER  225 

LOWRY,  O.  H.,  N.  J.  ROSEBROUGH,  A.  L.  FARR,  AND  R.  J.  RANDALL^  1951.     Protein  measure- 
ments with  the  Folin  phenol  reagent.   /.  Biol.  Chan.,  193:  265-275. 
LUMSDEN,  R.  D..  1975'.     Surface  ultrastructure  and  cytochemistry  of  parasitic  helminths.    Exp. 

Parasitol,  37 :  267-339. 

MACRAE,  E.  K.,   1967.     The  fine  structure  of  sensory  receptor  processes  in  the  auricular  epi- 
thelium of  the  planaria,  Dugesia  tigrina.  Z.  Zellforsch.  Mikrosk.  Anat.,  82:  479-494. 
MENTOX,  D.  N.,  AND  A.  Z.  EISEX,  1970.    The  structure  of  the  integument  of  the  sea  cucumber, 

Th\onc  briareits.  J.  Morfhnl.,  131:  17-36. 
NORREVANG,    A.,    1965.      Structure   and   function   of   the   tentacles    and    pinnules   of   Siboglinum 

ekmai    Jagersten     (Pogonophora)     with    special    references    to    the    feeding    problem. 

Sarsia.  21 :  37-47. 
POTSWALD,  H.  E.,   1971.     A  fine  structural  analysis  of  the  epidermis  and  cuticle  of  the  oligo- 

chaete,  Acolosoma  hengalense  Stephensen.  /.  MorphoL,  135:   185-212. 
PUTTER,  A.,  1908a.    Die  Ernahrung  der  Wassertier.  Z.  Allg.  Physiol.,  1 :  284-320. 
PUTTER,  A.,  1908b.    Die  Stoffhaus  Halt  des  Meeres.  Z,  Allg.  Physiol.,  1 :  321-368. 
REISH,  D.  J.,  AND  G.  C.  STEPHENS,  1969.     Uptake  of  organic  material  by  aquatic  invertebrates. 

V.   The  influences   of  age  on  the  uptake   of  glycine-C14   by  the   polychaete  Ncanthes 

arcnaceodcntata.   J\Iar.  Biol.,  3:  352-355. 
ROE,  P.,  1970.     The  nutrition  of  Pantnemertes  peregrina   (Rhynchocoela :  Hoplonermertea) .    I. 

Studies  on  food  and  feeding  behavior.    Biol.  Bull.,  139:  80-91. 

SIEGEL,  A.,  AND  E.  T.  DEGENS,  1966.     Concentration  of  dissolved  amino  acids  from  saline  wa- 
ters by  ligand-exchange  chromatography.  Science,  151  :  1097-1101. 
STEINMAN,  R.  M.,  1968.    An  electron  microscopic  study  of  ciliogenesis  in  developing  epidermis 

and  trachea  in  the  embryo  of  Xcnopus  Icvis.    Am.  J.  Anat..  122:  19-56. 
STEPHENS,  G.  C.,  1964.     Uptake  of  organic  material  by  aquatic  invertebrates.     III.  Uptake  of 

glycine  by  brackish  water  annelids.    Biol.  Bull.,  126:   150-162. 
STEPHENS,   G.   C,   1967.     Dissolved  organic  material  and  a  nutritional  source  for  marine  and 

estuarine  invertebrates.  Am.  Assoc.  Adv.  Sci.  Publ.,  83:  367-373. 
STEPHENS,  G.  C.,  1968.     Dissolved  organic  matter  as  a  potential  source  of  nutrition  for  marine 

animals.  Am.  Zoo/.,  8  :  95-106. 
STEPHENS,  G.  C,  AND  R.  A.  SCHINSKE,  1961.     Uptake  of  amino  acids  by  marine  invertebrates. 

Limnol.  Oceanogr.,  6:  175-181. 
TREVELYAN,  W.  E.,  P.  PROCTER,  AND  J.  S.  HARRISON,  1950.     Determination  of  sugars  on  paper 

chromatograms.   Nature,  166:  444-445. 

VON  BRAND,  T.,  1966.    Biochemistry  of  parasites.   Academic  Press,  New  York,  429  pp. 
WALSH,  G.  E.,  1965a.     Studies  on  dissolved  carbohydrates  in  Cape  Cod  waters.     I.  General 

survey.  Limnol.  Oceanogr.,  10 :  570-576. 
WALSH,  G.  E.,  1965b.     Studies  on  dissolved  carbohydrate  in  Cape  Cod  waters.     II.  Diurnal 

fluctuation  in  Oyster  Pond.   Limnol.  Oceanogr. ,  10 :  577-582. 
\YALSH,  G.  E.,   1966.     Studies  on  dissolved  carbohydrate  in  Cape  Cod  waters.     III.   Seasonal 

variation  in  Oyster  Pond  and  Wequaquet  Lake,  Massachusetts.    Limnol.  Oceanogr.,  11: 

249-256. 
WALSH,  G.  E.,  AND  J.  J.  DOUGLASS,  1966.     Vertical  distribution  of  dissolved  carbohydrate  in 

the  Sargasso  Sea  of  Bermuda.   Limnol.  Oceanogr.,  11:  406-408. 
\YANGERSKY,  P.  J.,  1952.     Isolation  of  ascorbic  acid  rhamnosides  from  sea  water.  Science,  115: 

685. 
WANGERSKY,  P.  J.,  AND  R.  R.  L.  GUILLARD,  1960.     Low  molecular  weight  organic  base  from 

the  dinoflagellate  Amphidininm  cartcri.    Nature,  185:  689-690. 
WEBB,  K.  L.,  AND  R.  E.  JOHANNES,  1967.     Studies  of  the  release  of  dissolved  free  amino  acids 

by  marine  zooplankton.    Limnol.  Oceanogr.,  12  :  376-382. 
WILBRANDT,  W.,  AND  T.  ROSENBERG,  1961.    The  concept  of  carrier  transport  and  its  corollaries 

in  pharmacology.    Pharmacol.  Rev.,  13  :   109-132. 
WILSOX,  R.  A.,  1969.     Fine  structure  of  the  tegument  of  the  miracidium  of  Fasciola  hcpatica 

L.  /.  Parasitol.,  55  :   124-133. 


Inference  :   B'wl.  Bull.,  154  :  226-240.     (April,  1978) 


ANTENNULAR  CHEMOSENSITIVITY  IN  THE  SPINY  LOBSTER, 

PANULIRUS  ARGUS:  STUDIES  OF  TAURINE 

SENSITIVE  RECEPTORS  x 

ZOLTAN  M.  FUZESSERY,2  WILLIAM  E.  S.  CARR  3  AND  BARRY  W.  ACHE 

Department  of  Biological  Sciences,  Florida  Atlantic  University,  Boca  Raton,  Florida  33431; 
and  C.  V  .  Whitney  Laboratory  for  Experimental  Marine  Biology  and  Medicine,  Rt.  1, 

Box  121,  St.  Augustine,  Florida  32084 

A  recent  study  of  the  antennular  chemosensory  system  in  Pamilirus  argus 
showed  that  the  low  molecular  weight  fractions  of  extracts  of  several  potential 
food  organisms  duplicated  the  receptor  activity  elicited  by  the  total  unfractionated 
extracts  (Ache,  Fuzessery  and  Carr,  1976).  Further,  in  at  least  one  of  the  above 
extracts,  the  amino  acids  were  shown  to  account  for  a  large  portion  of  the  activity 
with  taurine  being  the  single  most  stimulatory  amino  acid  (Johnson  and  Ache, 
1978).  Taurine  emerges  as  an  effective  stimulant  in  other  crustacean  studies  as 
well  (Case,  1964;  Crisp,  1967;  Ache,  1972;  Shepheard,  1974;  Carr  and  Gurin, 
1975;  Fuzessery  and  Childress,  1975;  Allison  and  Dorsett,  1977).  Taurine  sensi- 
tive receptors,  with  response  thresholds  as  low  as  10~10  M,  occur  on  both  the  lateral 
and  medial  antennular  filaments  of  the  spiny  lobster  (Fuzessery,  in  preparation). 
The  present  study  examines  the  molecular  specificity  of  taurine  sensitive  receptors 
by  comparing  the  stimulatory  capacity  of  taurine  with  that  of  taurine  analogs, 
derivatives,  and  structurally  related  compounds.  The  results  indicate  that  anten- 
nular taurine  receptors  of  P.  argus  are  characterized  by  a  narrow  and  consistent 
specificity  similar  to  that  of  the  taurine  endoreceptors  of  diverse  organisms. 

MATERIALS  AND  METHODS 

Excised  antennular  filaments  were  fitted  with  a  Sylgard  sleeve  over  their  proxi- 
mal end,  and  inserted  into  a  tubular  stimulating  chamber.  The  sleeve  separated 
fluid  in  the  stimulating  chamber  from  a  second  compartment  containing  about  10 
ml  of  Pamilirus  saline  (Mulloney  and  Selverston,  1974)  into  which  the  filament's 
proximal  end  projected.  The  preparation  was  perfused  with  oxygenated  Pannlints 
saline  introduced  under  pressure  through  a  tapered  glass  capillary  inserted  in  the 
cut  distal  tip  of  the  filament.  Axons  were  exposed  for  recording  by  cutting  the 
articular  membrane  between  the  fourth  and  fifth  most  proximal  segments  of  the 
filament  and  removing  the  cuticle  in  the  manner  of  removing  insulation  from  a 
wire.  Care  was  taken  to  place  minimal  stress  on  the  axon  bundle  during  this 
process.  Receptor  activity  was  recorded  extracellularly  using  a  monopolar  plati- 


study  was  supported  in  part  by  NSF  Grant  No.  PCM-73-07076  A02  (WESC)  and  a 
grant  from  the  Whitehall  Foundation  (BWA). 

2  Present  address  :   Neural  and  Behavioral  Biology  Program,  University  of  Illinois,  Urbana, 
Illinois  61801. 

3  Send   reprint   requests   to   W.    Carr,   Whitney   Marine   Laboratory,    Rt.    1,    Box    121,    St. 
Augustine,  Florida  32084. 

226 


TAURINE  RECEPTORS  IN  PANULIRUS 


227 


O 


C£ 

I— 

LU 

u 

z 
O 
u 


10 


0 


10 


2  4  6  8 

TIME   (SEC) 
FIGURE  1.    Temporal  profile  of  50-/J.I  stimulant  pulse  as  monitored  by  densitometry. 

num-iridium  hook  electrode  referenced  against  an  Ag-AgCl  pellet  submerged  in 
the  10  ml  saline  bath.  Signal  amplification  and  display  involved  standard  electro- 
physiological  instrumentation.  All  activity  was  stored  on  magnetic  tape  for  sub- 
sequent analysis. 

Reagent-grade  artificial  sea  water  (ASW,  MBL  formula)  continuously  entered 
the  stimulating  chamber  at  the  filament's  proximal  end  and  flowed  distally  over 
the  filament  at  a  rate  of  10  ml/min.  Fifty-microliter  pulses  of  test  stimulant  were 
pipetted  into  this  carrier  flow  of  ASW  through  a  port  2  cm  upstream  from  the 
preparation.  Figure  1  shows  the  temporal  profile  of  a  stimulus  pulse,  measured  by 
monitoring  a  pulse  of  methylene  blue  with  a  densitometer  located  at  the  midpoint  of 
the  tubular  compartment. 

All  compounds  employed  in  the  study  were  obtained  from  commercial  sources 
and  used  to  prepare  10~4  M  stock  solutions  in  ASW.  These  were  frozen  until 
needed,  thawed  and  serially  diluted  with  ASW  to  the  required  test  concentrations. 
All  solutions  were  tested  at  the  pH  (7.5)  and  temperature  (or.  22°  C)  of  the 
carrier  ASW  flow. 

The  general  protocol  in  each  experiment  was  to  search  for  single  taurine  sensitive 
neurons  while  stimulating  with  10~5  M  taurine.  Nerve  bundles  containing  taurine 
sensitive  units  were  sub-divided  until  only  the  taurine  sensitive  unit  remained,  or 
the  taurine  sensitive  unit  could  be  clearly  discriminated  from  background  multiunit 
activity.  Single  units  were  identified  as  such  by  consistent  amplitude,  configura- 
tion, regularity  of  interspike  interval  and  relative  response  latency.  Unless  other- 
wise indicated,  the  entire  group  of  compounds  tested  in  an  experiment  was  applied 
to  each  taurine  sensitive  receptor.  Taurine  was  applied  at  the  beginning,  midpoint 
and  end  of  each  test  series.  Any  loss  of  activity  in  response  to  the  final  taurine 
application  voided  that  test  series.  The  application  sequence  of  test  compounds  was 
randomized.  A  30-sec  period  followed  the  introduction  of  each  test  solution,  dur- 
ing which  time  the  filament  was  flushed  vigorously  with  two  1-ml  injections  of 
ASW.  Preliminary  trials  indicated  30  sec  was  sufficient  time  for  full  receptor  re- 
covery at  the  stimulant  concentrations  used.  Procedural  details  unique  to  specific 
experiments  are  included  in  Results. 


228 


FUZESSERY,  CARR,  AND  ACHE 


Response  parameters  of  maxinium  impulse  frequency,  number  of  impulses/ 
response  and  response  duration  were  quantified  by  playing  taped  responses  through 
a  window  discriminator  and  electronic  counter  (Haer  7400  series).  The  trans- 
formed output  was  displayed  on  a  storage  oscilloscope  in  the  form  of  a  post-stimu- 
lus time  histogram  of  the  impulses/100  msec  over  the  duration  of  the  response. 
Maximum  impulse  frequency  was  determined  by  observing  the  greatest  number  of 
impulses  collected  in  a  single  100  msec  time  interval.  The  number  of  impulses/ 
response  was  determined  as  the  sum  of  the  impulses  in  all  time  intervals  over  the 
duration  of  the  response.  The  index  of  relative  stimulatory  capacity  (RSC)  used 
in  this  study  to  compare  stimulants  was  calculated  as  the  number  of  impulses/ 
response  elicited  by  a  given  compound  divided  by  the  number  of  impulses/response 
elicited  by  taurine  X  100.  Hence,  the  RSC  value  for  taurine  on  each  receptor  is 
100.  In  a  few  cases  where  chemoreceptors  were  spontaneously  active,  an  index  of 
average  baseline  activity  in  the  absence  of  chemical  stimulation  was  calculated  and 
subtracted  from  the  activity  elicited  by  test  compounds  in  that  individual  receptor. 

RESULTS 
Preliminary  tests  of  tanrine — dose/response  relationships 

Taurine  was  tested  over  a  concentration  range  of  10'11  to  10~4  M  on  18  lateral 
and  18  medial  filament  receptors.  The  average  values  for  maximum  impulse  fre- 
quency, impulses/response  and  response  duration  are  shown  in  Figure  2.  Maxi- 


100- 


80- 


z 
o 

Q- 

- 


40- 


20- 


IMP/SEC 

(MAX) 


IMP/ RESPONSE 


DURATION 


CONCENTRATION    (-LOG        M) 


FIGURE  2.    Average  dose/response  relationships  given  by  36  antennular  receptors  to  taurine 
stimulation.     Ordinate  indicates  percentage  of  maximum  response,  i.e.,  that  to  10~4  M  taurine. 


TAURINE  RECEPTORS  IN  P.-lM'l.lKUS 


229 


cs 


51 

a  ^ 
S  J^ 


a?  E 


so 

a 


<o 

s» 


a   s 


to  '« 

•>  %•$ 

is 

ci    "<k 

>cq 


1  8 

§  S 


e 

S 


JS  -0 

• 


3     V 


to 
.» 


"C 
0 

i 
i 

^ 

-H      1\  -H      -H  -H 

# 

1 

rvj    r/} 

IO     Tt< 

Ol                 Tf     ~^                 O]      — 

- 

N 

O   lO 
"+    Ol 

01 

^H 

C 
(N 

^f  01 

0\  t- 

rr^                                 \rt 

-f                                           Ol 

2 

0  -J< 
GO   l^ 

0           -t 

oc 

M;    —  i 

3 

- 

t^1".       \^ 

-f    -t 

^H 
01 

2 

^^ 

i-H      —  H 

2 

10   U, 

3 

KS 

- 

•^5    OO 

<y 

n 

2 

S5 

c 

IH 

O 

- 

lO    -f 

a 
u 
tj 
u 

o 

C   01 

a 

'""""*      ^^ 

oc 

^  r^i 

- 

^H      T—  ( 

« 

^^ 

« 

0- 

* 

O\  -H 
-f    10 

tf> 

OS     -H 

Tf  l~- 

f4 

01  Ol 

00  00 

-H 

vO  0 
sC  >O 

•1 

3 

3 
3 

T"1                                                    "^ 

%  -o 

Vi 

3 
J 

g 

l"!!1^!  «| 

B    O 
J=    03 

II 

•i  B  2  1  -s  2  1  = 

E  -S_'S   =   7)  T3   7  ^ 

01            <            U    I            W 

-^  ii 

i 

O 

u 

rt 

O 
*^ 

i) 
f. 

c 
o 
a. 


o 
•/) 
£ 
i_ 

^ 


TD 
i) 

•u 


in 

nj 

o 

N 
O 
> 

U 

co 


c 

ct 


O 


O 
U, 


03 

U 

CO 


2 

o 


2M)  FUZESSERY,  CARR,  AND  ACHE 

mum  impulse  frequency  began  to  plateau  at  concentrations  of  10~5  and  10'*  M,  while 
the  total  number  of  impulses  and  the  response  duration  increased  regularly  over 
the  entire  concentration  range.  The  large  standard  deviations  of  each  parameter 
reflect  in  part  variations  in  sensitivity  among  receptors.  Individual  threshold  con- 
centrations ranged  from  10's  to  10  10  M.  As  subsequent  data  will  indicate  these 
deviations  also  reflect  variations  in  the  slopes  of  the  dose/response  curves  that  are 
characteristic  of  individual  receptors.  Based  on  these  findings,  a  standard  test 
concentration  of  10~5  M  taurine  was  chosen  to  insure  a  strong  yet  nonsaturating 
response  from  all  receptors. 

Specificity  of  taurine  sensitive  receptors 

The  stimulatory  capacity  of  taurine  (==  2-aminoethyl  sulfonic  acid)  was  com- 
pared with  that  of  three  analogs  and  five  related  sulfonic  acids.  All  compounds 
were  tested  at  10~5  M  on  each  of  21  taurine  sensitive  receptors  on  the  lateral  and 
medial  filaments.  Calculations  of  the  relative  stimulatory  capacity  (RSC)  for  each 
compound  on  each  receptor  are  summarized  in  Table  I.  A  comparison  of  the  RSC 
values  reveals  that  only  taurine  and  its  carboxylic  and  sulfinic  acid  analogs,  /?- 
alanine  and  hypotaurine,  stimulated  all  receptors. 

Other  comments  on  data  in  Table  I  are  presented  below  following  the  presenta- 
tion of  some  additional  results. 

To  further  define  response  specificity,  thirteen  additional  compounds,  struc- 
turally-related to  taurine,  were  tested  at  10~n  M  on  each  of  18  taurine  sensitive 
lateral-  and  medial-filament  receptors.  The  resulting  RSC  values  are  presented  in 
Table  II.  Structural  formulae  of  compounds  that  were  tested  are  shown  in  Fig- 
ure 3.  Conclusions  concerning  receptor  specificity  are  summarized  below. 

1 .  Compounds  with  one  terminal  basic  group  and  one  terminal  acidic  group 
separated  by  two  carbon  atoms  were  most  effective.     Taurine  and  its  analogs, 
hypotaurine,  /3-alanine  and  2-aminoethyl-phosphonic  acid  all  meet  these  structural 
requirements.     Though  less  stimulatory  than  taurine,  the  analogs  hypotaurine  and 
/3-alanine  stimulated  all  receptors  and  their  RSC  values  with  individual  receptors 
were   consistently   similar.      The    phosphonic   acid    analog   was    dramatically    less 
effective  and  elicited  a  response  from  only  one  of  the  21  receptors  tested  (Table  I). 

2.  Compounds  with  one  terminal  basic  group  and  one  terminal  acidic  group 
separated  by  more  than  two  carbon  atoms  were  also  effective  although  the  RSC 
values  decreased  with  the  distance  of  separation  of  the  charged  groups.     This  is 
illustrated  in  Table  II  by  comparing  the  RSC  values  of  the  following:   /3-alanine  > 
y-amino-n-butyric   acid    (GAB A)  >  5-aminovaleric    acid    >    6-aminocaproic   acid. 
Note  also  that  two  isomers  of  GABA,  2-aminobutyric  acid  and  3-aminobutyric  acid, 
with  nonterminal  amine  groups,  are  markedly  less  effective  than  GABA. 

3.  Compounds  with  a  terminal  basic  group  and  a  terminal  acidic  group  separ- 
ated by  only  one  carbon  atom  (rather  than  two  carbon  atoms)  were  markedly  less 
effective.    This  is  shown  in  Table  I  by  the  low  incidence  of  receptor  stimulation  and 
the  low  RSC  value  of  aminomethyl  sulfonic  acid  (AMS).     Note  that  AMS,  like 
taurine,  has  terminal  amine  and  sulfonic  acid  groups.    Later  in  the  report  data  are 
presented  to  show  that  the  AMS  analog,  glycine,  as  well  as  other  a-amino  acids 
are  virtually  ineffective  in  taurine  sensitive  receptors. 


TAURINE  RECEPTORS  IN  PANULIRUS 


TAURINE    AND  ANALOGS 


H2N-CH2-CH2-S03H 
Taurine  (VA) 


H2N-CH2-CH2-S02H 
Hypotaurine  (VA) 


H2N-CH2-CH2-C02H 
^-Alanine  (VA) 


H2N-CH2-CH2-P03H2 
2-Aminoethylphosphonic  acid  (I) 


OTHER   SULFONIC  ACIDS 


COMPOUNDS  WITH   TERMINAL 
BASIC  AND   ACIDIC  GROUPS 


H2N-CH2-S03H 
Aminomethyl  sulfonic  acid  (SA) 

CH3-CH2-S03H 
Ethane  sulfonic  acid  (I) 

HO-CH2-CH2-S03H 
Hydroxyethane  sulfonic  acid  (I) 

CI-CH2-CH2-S03H 
2-Chloroethane  sulfonic  acid  (I) 

HO,C-CH-CH,-SO,H 

C.          \  £.3 

NH2 
Cysteic  acid  (I) 

COMPOUNDS  WITH  NON-TERMINAL 
BASIC     GROUPS 

CHv-ChL-CH-CCvH 

O        e.    |  £. 

NH2 
2-Aminobutyric    acid  (I) 

CH,-CH-CH,-CO,H 

O      |  C.  C. 

NH2 
3-Aminobutyric  acid  (I) 

H2N-CH2-CH-C02H 
NH2 

2,3-Diaminopropionic  acid  (I) 

H2N-CH2-CH2-CH-C02H 
NH2 

2,4-Aminobutyric  acid  (I) 

R-CH-C02H 

NH2 
cx-Amino  acids  (I) 

FIGURE  3.  Structural  formulae  of  compounds  tested  on  taurine  sensitive  receptors.  Indices 
of  relative  activity  are  as  follows:  (YA),  very  active;  (A),  active;  (SA),  slightly  active;  (I), 
virtually  inactive. 


H2N-CH2-C02H 
Glycine  (I) 

H2N-CH2-CH2-CH2-C02H 
l-Aminobutyric   acid  (A) 

H2N-(CH2)4-C02H 
5-Aminovaleric  acid   (A) 

H2N-(CH2)5-C02H 
6-Aminocaproic  acid  (SA) 

H~N-CH,-CH-CH,-CO,H 

C.  £.     |  C.  C. 

OH 
l-Amino-/9-hydroxybutyric  acid  (SA) 

H,N-CH,-CH-COC)H 
Z  £.   i  f. 

CH3 
/fl-Aminoisobutyric  acid  (A) 

H,N-C-NH-CH,-CO,H 

2        M  22 

NH 

Guanidoacetic  acid   (I) 

H,N-C-NH-CH0-CH,,-CO,H 
2      I,  222 

NH 
Guanidopropionic    acid  (I) 

H,N-CH,-CH,-C-NH-CH-CO,H 

2  2         2  ii  i  2 


0 


CH, 


/fl-Alanylalanine  (I) 


H2N-CH2-CH2-C-NH-CH2-C02H 


/9  -Alanylglycine  (I) 


232 


FUZESSKRY,  CARR,  AND  ACHE 


s 

x 


II 


-C) 


""  -5 

S    S 


S    <* 

<s    K. 


i   e 

s   <>, 

"3  ^S 
" 


w 

CQ      ^ 


e    a 


- 

a 


"Xi 
nS 

.          tS) 

- 


•1 

3 
99 

O  t~»          c-1  10  ON          >O  r—  -rt<  PO  vO 

_U-U          OJ-U-U          _U-U4J4J4J 
Till            TITlTl            TlllTITlTl 

•> 

l> 

, 
<j 

O   >O           »-H 

- 

^O   ^2          ^O   *•&   ^J          ^~^   ^J   *~^           ^^ 

2 

^s   ^^           ^-H  r^5   O"\          ^^  ^^^          O^ 
oo  oo         rs  -H  -H 

S 

t"""*  o^          ^^  ^*  ^^          <^^  r*^  QQ  ^^5 

t^-     t^»                   TO      OJ      rvj                   ^H 

S 

O  -f          O\  ON  QC          i/:  *-t 
^O  ^i          C^l 

2 

O  -f         oo  r<i  oo         tN  oo 

2 

f^    ^*             t~^-    f^    c^i             LO 

r-~  \O          ^^1   ^"~f   '•^ 

u 

- 

"•O     ^H             IO    IT) 

\O   iO 

1 

c 

c 

O   "f                 ^O 
IO    -t1 

o 
a 

* 

'•H   ON           C^J 
Tf   IO           ^H 

(3 

oo 

00    '-i           CN 

IO     1^                 -H 

*• 

Ov    1^ 

•o 

o  >o       oo 

,0 

IO    CN 

- 

OO   00 

„ 

00  CO 

tv}    ^^ 

« 

00  t^ 

00   NO 

- 

£3 

•c 
c 

c 
E 

I    "i<|^l|,  |g.i  §•.!•!!.£•§! 
=  c  °^'s  o  J  g  8  §  s^  j^^^'g.s 

2'?    c_.    =    C<    t,    c-u<    =<j?*ppc«    p 

B.J"s-B*s's  f>,'s'Sr?  3^  &J  «5< 

>.<<     «<<-J-C<     3JOJ^<<r^4 

nj 

H 

o 

-4— ' 

vl) 
-*-» 

o 


OJ 

at 


4) 

"co 

U 

en 
Pi 

0) 

o/. 

U 


TAURINE  RECEPTORS  IX  r.lXCLIRUS  233 

4.  Taurine  (U'rivutives  lacking  the  basic  ainine  group  were  markedly  less  effec- 
tive.    This  is  shown  by  the  low  incidence  of  receptor  stimulation  and  the  low  RSC 
values  of  ethane  sulfonic  acid,  hydroxyethane  sulfonic  acid  and  chloroethane  sul- 
fonic  acid  (Table  I). 

5.  The  addition  of  a  neutral  side  chain  decreased  the  effectiveness  of  a  com- 
pound.    y-Amino-/3-hydroxybutyric  acid  differs  from  GABA  by  having  a  hydroxyl 
group  and  yet  has  a  much  lower  RSC  value   (Table  II).     Likewise,  /2-aminoiso- 
butyric  acid  differs  from  /2-alanine  by  having  a  methyl  group  and  yet  has  a  much 
lower  RSC  value. 

6.  Compounds   with   an   alplia-nmlne   group   in   addition   to   a   terminal   amine 
group  were  virtually  ineffective.     Note  in  Table  II  that  2,3-diaminoproprionic  acid 
and  2,4-aminobutyric  acid  are  virtuallv  inactive,  whereas  the  closely  related  com- 
pounds, /3-alanine  and  GABA,  have  marked  activities. 

7.  Compounds  in  which  the  terminal  basic  group  is  a  guanido  group  rather 
than  an  amine  group  were  far  less  effective.    This  is  shown  in  Table  II  by  the  very 
low  RSC  values  of  guanidoacetic  and  /3-guanidopropionic  acid. 

8.  Two  dipeptides  containing  the  stimulatory  amino  acid  /3-alanine  were  in- 
effective thereby  suggesting  that  activity  is  lost  when  the  carboxyl  group  is  in- 
volved in  a  peptide  bond  (Table  II).     Likewise,  the  presence  of  two  acidic  groups 
apparently  negates  activity  as  shown  by  the  ineffectiveness  of  cysteic  acid  (Table  I ). 

The  data  in  Table  II  also  indicate  the  existence  of  a  distinct  relationship  be- 
tween the  average  RSC  value  of  a  compound  and  the  number  of  receptors  respond- 
ing to  that  compound.  Hence  compounds  with  higher  RSC  values  elicited  responses 
from  a  larger  percentage  of  the  receptors.  This  relationship  implies  strongly  that 
the  "taurine  receptors"  have  a  consistent  and  predictable  specificity  and  thus  appear 
to  comprise  a  distinct  receptor  class.  Regarding  this  specificity,  no  (differences 
were  observed  between  taurine  sensitive  receptors  present  on  the  lateral  or  the 
medial  antennular  filaments. 

Additional  tests  of  receptor  specificity 

In  order  to  gain  further  insight  into  the  restricted  specificity  of  these  cells,  12 
a-amino  acids,  3  organic  acids,  and  the  quaternary  amine,  glycine  betaine,  were 
tested  at  a  concentration  of  10~5  M  on  additional  taurine  sensitive  receptors.  Table 
III  shows  that  individual  compounds  were  applied  to  5  to  65  receptors  and  that 
none  of  the  new  compounds  cited  above  elicited  responses.  As  in  Tables  I  and  II. 
the  taurine  analogs  included  in  this  test-series  stimulated  all  receptors,  whereas 
GABA  and  /3-aminoisobutyric  acid  stimulated  a  large  percentage  of  them.  RSC 
values  were  not  computed  because  in  this  phase  of  the  study  all  of  the  compounds 
were  not  tested  on  all  of  the  receptors.  The  inability  of  all  a-amino  acids  to  activate 
taurine  sensitive  receptors  strongly  supports  the  preceding  results  which  indicated 
that  amine  groups  in  the  alpha  position  reduced  effectiveness.  The  ineffectiveness 
of  the  organic  acids  tested  supports  the  earlier  conclusion  that  stimulatory  mole- 
cules require  both  positively  and  negatively  charged  atoms. 

Quantitative  effects  of  stimulatory  compounds 

Whereas  the  RSC  values  presented  earlier  for  various  compounds  showed  a 
consistent  ranking  with  individual  receptors,  considerable  variations  were  apparent 


234 


FUZESSERY,  CARR,  AND  ACHE 


CELL  A 
120- 


80 


5    40 


',( )i  i 


200' 


18' 


12' 


z 
o 


65432 


7        6 


432 


CONCENTRATION  (-LOG]QM) 


3OO' 


1OO 


18' 


12' 


FIGURE  4.  Three  concentration-dependent  response  parameters  (maximum  impulse  fre- 
quency, total  impulses,  and  response  duration)  of  two  receptors  stimulated  with  taurine  (T), 
/3-alanine  (A),  hypotaurine  (H)  and  7-amino-n-butyric  (G). 

in  the  RSC  values  obtained  with  individual  compounds  on  different  receptors.  For 
example,  /3-alanine  was  usually  the  second  more  stimulatory  compound,  yet  its 
RSC  values  ranged  from  5  to  94  (Table  I).  To  explore  the  basis  of  this  variability, 
a  detailed  evaluation  was  made  of  three  response  parameters  of  two  receptors 
tested  with  graded  concentrations  of  the  four  most  stimulatory  compounds 
(Fig.  4).  In  both  receptors,  the  concentration  functions  of  the  three  parame- 
ters described  a  series  of  roughly  parallel  curves.  Also  in  both  receptors,  maximum 
impulse  frequency  reached  a  maximum  value  and  did  not  increase  at  higher  con- 
centrations. In  the  slower  adapting  receptor  (Fig.  4A),  impulses/response  and 
response  duration  continued  to  increase  with  concentration ;  while  in  the  more 
rapidly  adapting  receptor  (Fig.  4B),  these  parameters  reached  maximum  values  at 
approximately  the  same  concentration  as  frequency.  This  variation  between  the 
slow  and  fast  adapting  receptors  likely  results  from  the  mode  of  stimulus  introduc- 
tion which  was  a  pulse  with  an  exponential  dilution  profile  (see  Fig.  1).  Hence, 
as  concentration  increased,  the  period  during  which  the  pulse  remained  at  a  supra- 
threshold  concentration  also  increased,  thereby  prolonging  the  response  of  the 
slow-adapting  receptor. 


TAURINE  RECEPTORS  IN  PANULIRUS 


235 


In  addition  to  the  variations  cited  above,  individual  receptors  also  varied  in 
sensitivity  and  in  the  profile  of  their  dose-response  curves.  Note  that  the  concen- 
tration function  of  impulses/response  rises  more  sharply  in  the  receptor  represented 
in  Figure  4B  than  that  in  Figure  4A.  Individual  variations  in  sensitivity  indicate 
that  response  to  the  standard  test  concentration  ( 10~5  M)  will  not  occupy  the  same 
relative  position  on  the  dose-response  curve  of  each  receptor.  In  less  sensitive 
receptors,  a  1O5  M  concentration  of  a  given  compound  may  be  close  to  the  threshold 
concentration.  In  very  sensitive  receptors,  a  10~5  M  concentration  may  be  close  to 
the  plateau,  concentration.  To  return  to  the  example  of  variation  in  the  individual 
RSC  values  of  /3-alanine,  it  can  be  inferred  that  in  a  very  sensitive,  rapidly  adapt- 
ing receptor,  the  test  concentration  of  10~5  M  may  be  near  the  plateau  concentrations 
of  both  taurine  and  /J-alanine,  resulting  in  approximately  equal  RSC  values.  Con- 
versely, in  a  less  sensitive  receptor,  the  test  concentration  may  be  near  the  threshold 
concentration  of  /3-alanine,  resulting  in  a  very  low  RSC  value.  This  inherent  vari- 
ability among  receptors  underscores  the  necessity  of  comparing  RSC  values  only 
in  cases  where  all  compounds  are  applied  to  each  receptor  in  the  test  population. 
These  factors  may  also  explain  why  GAB  A,  the  fourth  most  stimulatory  com- 
pound, did  not  activate  all  receptors  (Tables  II  and  III).  In  less  sensitive  recep- 
tors, the  test  concentration  may  be  below  the  threshold  concentrations  for  GABA 
(see  also  Fig.  4). 

That  the  three  response  parameters  detailed  in  Figure  4A,  B  describe  a  series  of 
roughly  parallel  curves  suggests  that  these  compounds  effect  impulse  generation  in 


TABLE  III 

Sensitivity  of  taurine  sensitive  receptors  to  taurine  analogs,  a-amino  acids  and  other  compounds, 
compounds  were  tested  at  10~5  M. 


All 


Compound 

Number  of  receptors 
tested 

Number  of  receptors 
activated 

Receptors  activated 
(%) 

L-or-Alanine 

43 

0 

— 

/3-Alanine 

33 

33 

100 

a-Aminoisobutyric  acid 

8 

0 

— 

/3-Aminoisobutyiic  acid 

8 

5 

63 

•}-Amino-n-butyric  acid 

15 

14 

93 

L-Aspartic  Acid 

30 

0 

— 

Citric  acid 

10 

0 

— 

L-Glutamic  acid 

19 

0 

•  — 

Glycine 

65 

0 

— 

Glycine  betaine 

43 

0 

— 

Hydroxy-L-proline 

7 

0 

— 

Hypotaurine 

9 

9 

100 

L-Isoleucine 

7 

0 

— 

L-Leucine 

12 

0 

— 

L-Lysine 

8 

0 

— 

Propionic  acid 

19 

0 

— 

Succinic  acid 

19 

0 

— 

Taurine 

65 

65 

100 

L-Tryptophan 

7 

0 

— 

L-Tyrosine 

7 

0 

— 

L-Valine 

5 

0 

— 

236 


FUZESSERY,  CARR,  AND  ACHE 


TAU    10"5M 


B  ALA    10"4M 


H-TAU    10"4,M 


GABA    1CT3M 


SEC 


FIGURE  5.  Response  of  single  receptor  to  four  stimulants  with  concentrations  adjusted  to 
elicit  essentially  equal-intensity  responses.  Time  bar  is  1  sec.  Tau  represents  taurine;  /3-ala, 
/3-alanine;  H-Tau,  hypotaurine;  and  GABA,  7-amino-n-butyric  acid. 


a  manner  mimicking  the  concentration  function  of  a  single  compound.  A  less 
stimulatory  compound  effects  receptor  response  in  the  same  manner  as  a  more 
stimulatory  compound  applied  at  a  lower  concentration.  The  functional  implica- 
tion is  that  the  receptor  response  elicited  by  taurine  at  10~5  M  would  be  very  similar 
to  that  of  /3-ala  at  1CH  M  and  GABA  at  10'3  M,  as  suggested  by  Figure  5. 

DISCUSSION 

These  results  indicate  that  the  antennules  of  the  spiny  lobster,  Paniilinis  argns, 
possess  taurine  sensitive  receptors  with  a  narrow  and  consistent  response  specificity. 
Previous  electrophysiological  studies  of  crustacean  chemoreception  indicate  that 
individual  receptors  exhibit  differential  specificities  to  amino  acids  and  related  com- 
pounds (Laverack,  1964;  Case,  1964;  Ache,  1972;  Shepheard,  1974;  Fuzessery 
and  Childress,  1975).  The  two  studies  dealing  most  thoroughly  with  receptor 
specificity  (Case,  1964;  Shepheard.  1974)  provided  values  of  the  relative  activity 
of  compounds  obtained  by  pooling  results  from  the  entire  population  of  receptors 
tested.  This  practice  treats  all  receptors  as  being  effectively  monotypic  with  re- 
spect to  specificity.  Moreover,  in  the  above  studies  all  compounds  were  not  ap- 
plied to  each  receptor  in  the  test  population.  The  latter  procedure  is  essential  for 
an  analysis  of  both  the  specificity  and  the  inherent  variability  of  individual  recep- 
tors. However,  regarding  the  taurine  sensitive  receptors  analyzed  in  the  current 
study,  some  corroborating  evidence  is  present  in  the  study  by  Shepheard  (1974) 
on  another  decapod  crustacean,  Honiarns  aincricanus.  In  a  case  in  which  43 


TAURIXE  RECEPTORS  IX  PAXL'LIRUS  237 

amino  acids  and  related  compounds  were  applied  to  a  single  receptor,  only  taurine 
and  /3-alanine  were  stimulatory. 

It  is  important  to  emphasize  that  our  current  documentation  of  the  distinct 
specificity  of  taurine  receptors  in  P.  art/us  was  made  possible  largely  by  our  early 
recognition  of  the  extreme  sensitivity  of  these  receptors  to  taurine.  This  recogni- 
tion led  to  our  decision  to  work  with  a  dilute  (10~5  M)  standard  test  concentration. 
As  shown  clearly  in  Figure  4,  the  apparent  specificity  of  a  receptor  becomes  less 
distinct  as  the  test  concentration  is  increased.  The  failure  of  earlier  workers  to 
detect  receptor  classes  with  distinct  specificity  in  crustaceans  may  be  due  to  using 
high  stimulant  concentrations  (ca.  10~2  M). 

Taurine  sensitive  receptors  with  a  somewhat  similar  specificity  to  those  found 
in  antennules  of  Panulinis  arc/us  have  been  reported  in  endoreceptors  serving  a 
variety  of  functions.  In  the  examples  cited  below,  note  that  the  activity  of  taurine 
was  mimicked  by  the  analogs  hypotaurine  and  /3-alanine  but.  in  the  instance  where 
tested,  not  by  the  phosphonic  acid  analog.  Also,  in  cases  where  tested,  the  taurine 
receptors  were  markedly  less  responsive  to  a-amino  acids.  Taurine  is  effective  in 
suppressing  induced  heart  seizures  in  dogs,  and  this  action  is  most  effectively 
mimicked  by  /?-alanine,  hypotaurine  and  GABA  but  not  by  glycine  or  a-alanine 
(Barbeau,  Tsukada,  and  Inoue.  1976).  Induced  arrhythmia  in  dogs  is  suppressed 
by  taurine  but  not  by  ethanesulfonic  acid  and  other  compounds  lacking  both  basic 
and  acidic  groups  (Welty,  Read  and  Byington.  1976).  In  an  active  transport  sys- 
tem in  human  blood  platelets,  taurine  uptake  is  inhibited  competitively  by  /2-alanine 
and  hypotaurine  but  not  by  the  phosphonic  acid  analog  (Grant  and  Nauss,  1976). 
Similarly,  taurine  uptake  by  rat  brain  slices  is  inhibited  competitively  by  hypotaurine 
and  /?-alanine  but  not  by  a-amino  acids  (Kaczmarek  and  Davison,  1972;  Lahdes- 
maki  and  Oja,  1973). 

The  inhibitory  effect  of  GABA  (=  4-aminobutyric  acid)  on  crustacean  stretch 
receptors  is  most  effectively  mimicked  by  3  and  5  carbon  chain  amino  acids  with 
terminal  amine  groups,  i.e..  /?-alanine  and  5-aminovaleric  acid.  6-Aminocaproic 
acid  was  less  effective,  and  glycine  was  essentially  without  effect.  Taurine  was  less 
effective  than  its  carboxylic  acid  analog,  /3-alanine  (Robbins,  1959;  Edwards  and 
Kuffler,  1959).  As  in  the  present  antennular  system,  the  latter  workers  reported 
that  the  addition  of  neutral  side  chains  reduced  effectiveness,  and  that  the  presence 
of  both  the  acidic  and  the  basic  groups  were  essential.  In  general,  the  GABA  system 
appears  to  resemble  the  present  one,  differing  primarily  in  that  the  ideal  separation 
of  opposite  charges  is  three,  rather  than  two,  carbon  atoms. 

Similarities  in  the  apparent  specificity  of  both  internal  and  external  taurine  re- 
ceptors lend  support  to  the  concept  that  systems  for  molecular  recognition,  once 
evolved,  may  be  preserved  and  used  in  a  variety  of  functions,  ranging  from  solute 
uptake  and  regulation  to  chemical  sensing,  synaptic  transmission  and  others  (Kit- 
tredge,  Takahashi.  Lindsey,  and  Lasker,  1974;  Lenhoff,  1975).  In  the  future  we 
hope  to  provide  a  detailed  model  of  the  antennular  receptor  site  for  taurine.  How- 
ever, the  presentation  of  such  a  model  must  await  the  testing  of  several  additional 
analogs  and  derivatives  that  are  not  available  commercially  and  hence  must  be 
specially  synthesized. 

Pannlirns  argits  is  a  predator/scavenger  that  feeds  on  a  variety  of  molluscs, 
arthropods,  echinoderms  and  fish  (Herrnkind,  VanDerwalker  and  Barr,  1975). 


238  FUZESSERY,  CARR,  AND  ACHE 

Analyses  of  tissue  extracts  of  marine  molluscs,  arthropods,  echinoderms  and  fish 
show  that  the  taurine  concentration  ranks  from  first  to  fifth  in  the  total  pool  of 
free  amino  acids  (Carr,  1976;  Carr,  Blumenthal  and  Netherton,  1977).  Taurine 
is  certainly  the  most  abundant  /3-amino  acid  in  most  marine  animals.  According  to 
Awapara  (1976  p.  1),  "taurine  exists  uncombined  and  distributed  throughout  the 
animal  kingdom  in  a  manner  almost  unparalleled  by  any  known  small  organic  mole- 
cule." Therefore,  it  is  clear  that  taurine  receptors  could  be  expected  to  provide 
sensory  information  on  the  proximity  of  an  array  of  suitable  food  organisms.  Al- 
though the  taurine  analogs,  hypotaurine  and  /?-alanine,  are  also  very  stimulatory  to 
the  antennular  taurine  receptors,  both  of  these  compounds  occur  in  only  minor 
concentrations  in  the  tissues  of  most  organisms  (Sturman,  Hepner,  Hofmann,  and 
Thomas,  1976;  Awapara,  1976).  Hence,  one  must  assume  that  the  potential  chemo- 
sensory  role  of  these  other  stimulants  is  far  less  than  that  of  taurine. 

It  is  of  special  interest  that  taurine  receptors  are  very  insensitive  to  a-amino 
acids,  particularly  since  these  compounds  are  present  in  high  concentration  in  the 
tissues  of  many  marine  animals.  The  antennular  chemosensory  system  appears  to 
be  so  constructed  that  a  portion  of  the  total  receptor  population  responds  to  a 
single,  ubiquitous  /?-amino  acid,  and  is  functionally  insensitive  to  other  commonly 
occurring  amino  acids.  Perhaps  the  significance  of  this  finding  resides  in  the  fact 
that  a-amino  acids  are  common  constituents  of  sea  water,  occurring  at  individual 
concentrations  of  1O7  to  1O9  M  (Duursma,  1965).  Comparable  concentrations  of 
taurine  have  not  been  reported.  A  plausible  speculation  may  be  that  dissolved  a- 
amino  acids  produce  a  chemical  "white  noise"  against  which  chemosensory-based 
discrimination  must  occur.  Taurine  receptors  would  be  unaffected  by  ambient  a- 
amino  acid  levels,  and  therefore  may  provide  less  ambiguous  information  regard- 
ing the  proximity  of  potential  prey. 

In  the  present  antennular  system,  taurine  appears  to  comply  with  Beets'  (1971) 
definition  of  a  nonideal  mono-osmatic  odorant,  i.e.,  a  single  compound  which  acti- 
vates a  single  receptor  at  lower  concentrations  than  other  compounds  within  the 
specificity  of  that  receptor.  In  addition,  when  one  considers  the  chemical  composi- 
tion of  the  natural  foods  of  P.  argns,  taurine  is  the  only  compound  that  we  have 
tested  which  is  likely  to  be  present  in  sufficient  concentrations  to  activate  these 
receptors.  From  a  functional  standpoint,  the  antennular  taurine  receptors  can  be 
considered  specialist  receptors  which  may  serve  to  monitor  the  presence  of  a  single 
compound.  This  is  particularly  significant  in  that  it  is  one  of  the  few  cases  in  which 
specialist  receptors  have  been  identified  which  may  play  a  role  in  the  mediation  of 
feeding  behavior,  and  the  first  documentation  of  such  receptor  organization  in 
crustacean  chemoreceptors. 


SUMMARY 

1.  Taurine  sensitive  receptors  in  the  antennules  of  the  spiny  lobster,  Punulirns 
argus,  were  identified  electrophysiologically. 

2.  Recordings  from  single  receptors  revealed  a  narrow  and  consistent  specificity 
when  tested  with  taurine,  taurine  analogs  and  derivatives,  and  structurally  related 
compounds. 


TAURINE  RECEPTORS  IN  PANULIRUS  2.W 

3.  Taurinc  was  the  most  stimulatory  compound  tested.     Threshold  concentra- 
tions for  36  individual  receptors  ranged  from  10"8  to  10  10  M. 

4.  The  taurine  analogs,  hypotaurine  and  /2-alanine,  were  also  very  effective  but 
the  phosphonic  acid  analog  of  taurine  was  ineffective. 

5.  Regarding  receptor  specificity,  receptor  stimulation  was  greatest  with  com- 
pounds having  single  terminal  basic   (amine)  and  acidic  groups  separated  by  two 
carbon  atoms.     Compounds  having  terminal  basic  and  acidic  groups  separated  by 
three  to  five  carbon  atoms  were  also  active.     However,   activity   decreased  with 
the  distance  of  separation  of  charged  groups. 

6.  Alf>lia-am\no  acids  and  compounds  with   terminal   basic   and   acidic  groups 
separated  by  only  one  carbon  atom  were  virtually  ineffective. 

7.  Receptor  stimulation  was  markedly  less  with  structurally  related  compounds 
that  either  lacked  a  terminal  amine  group,  had  additional  amine  or  acidic  groups,  or 
had  neutral  side  chains. 

8.  Dose/response    relationships    of   four    differentially    stimulatory   compounds 
(taurine,  hypotaurine.  /?-alanine  and  y-aminobutyric  acid)  applied  to  single  recep- 
tors were  compared  and  found  to  describe  a  series  of  roughly  parallel  lines.     This 
implies  that  a  less  stimulatory  compound  effects  receptor  response  in  the  same  man- 
ner as  a  more  stimulatory  compound  applied  at  a  lower  concentration. 

9.  The  possible  role  of  taurine  in  food  finding,  and  the  similarity  of  the  speci- 
ficity of  antennular  taurine  receptors  and  taurine  endoreceptors  identified  in  various 
organisms  are  discussed. 

LITERATURE  CITED 

ACHE,  B.  W.,  1972.     Amino  acid  receptors  in  the  antennules  of  Homarus  aincricanus.    Coinp. 

Biochcm.  Physiol.,  42:  807-811. 
ACHE,  B.  W.,  Z.  M.  FUZESSERY,  AND  W.  E.  S.  CARR,  1976.    Antennular  chemosensitivity  in  the 

spiny  lobster,  Panulirus  argus:    comparative  tests  of  high  and  low  molecular  weight 

stimulants.    BioL  Bull,  151 :  273-282. 
ALLISON,   P.,   AND  D.  A.   DORSETT,    1977.     Behavioral   studies   on   chemoreception   in   Balanus 

hamcri.  Mar.  Bchav.  Physio!.,  4:  205-217. 
AWAPARA,  J.,  1976.     The  metabolism  of  taurine  in  the  animal.     Pages  1-21  in  R.  Huxtable  and 

A.  Barbeau,  Eds.,  Taurinc.   Raven  Press,  New  York. 
BARBEAU,  A.,  Y.  TSUKADA,  AND  N.  INOUE,  1976.    Neuropharmacologic  and  behavioral  effects  of 

taurine.     Pages  253-267  in  R.  Huxtable  and  A.  Barbeau,  Eds.,  Taurinc.    Raven  Press, 

New  York. 
BEETS,  M.  G.  J.,  1971.     Olfactory  response  and  molecular  structure.     Pages  257-321   in  L.  M. 

Beidler,  Ed.,  Handbook  of  sensory  physiology,  Vol.  IV ,  part  1.    Springer-Verlag,  New 

York. 
CARR,  W.  E.  S.,  1976.     Chemoreception  and  feeding  behavior  in  the  pigfish,  Ortliopristis  chry- 

soptcrus:    characterization   and   identification   of    stimulatory    substances    in    a    shrimp 

extract.     Coinp.  Biochcm.  Physiol..  55A :   153-157. 
CARR,  W.  E.  S.,  AND  S.  GURIN,   1975.     Chemoreception  in  the  shrimp,  Palaemonetes  pnqio : 

comparative  study  of  stimulatory  substances  in  human  serum.  Biol.  Bull.,  148:  380-392. 
CARR,  W.  E.  S.,  K.  M.  BLUMENTHAL,  AND  J.  C.  NETHERTON,  1977.     Chemoreception  in  the  pig- 
fish,  Ortliopristis  chrysoptcrns :   the  contribution  of  amino  acids  and  betaine  to  stimula- 
tion of  feeding  behavior  by  various  extracts.  Coinp.  Bioclicui.  Physiol.,  58A  :  69-73. 
CASE,  J.,   1964.     Properties  of  the  dactyl  chemoreceptors  of  Cancer  antcininrins  Stimpson  and 

C.  producing  Randall.  Biol.  Bull..  127  :  428-446. 

CRISP,  D.  J.,  1967.     Chemoreception  in  cirripedes.    Biol.  Bull.,  133  :   128-146. 
DUURSMA,  E.  K.,  1965.     The  dissolved  organic  constituents  of  sea  water.     Pages  433-473  in 

J.  P.  Riley  and  G.  Skirrow,  Eds.,  Chemical  oceanography.   Academic  Press,  New  York. 


240  FUZESSERY.  CARR,   AND  ACHE 

KHXVAKDS.  C,  AND  S.  W.  KI-KKLER,  1959.    The  blocking  HTect  of  •y-aminobutyric  acid   (GAliA) 

and  the  action  of  related  compounds  on  single  nerve  cells.  J.  Neurochem.,  4:   19-30. 
FUZESSERY,  Z.  M.,  AND  J.  J.  CHILDRESS,  1975.     Comparative  chemosensitivity  to  amino  acids 

and  their  role   in   the  feeding  activity   of  bathypelagic   and   littoral   crustaceans.    Bwl. 

Bull.,  149:  533-538. 
GANT,  Z.  N.,  AND  C.  B.  NAUSS,  1976.     Uptake  of  taurine  hy  human  blood  platelets:    a  possible 

model   of  the  brain.     Pages  99-121    in   R.   Huxtable  and  A.   Barbeau,   Eds.,    Taurine. 

Raven  Press,  New  York. 
HERRNKIND,  W.,  J.  VANDERWALKKR,   AND  L.   BARR,   1975.     Population  dynamics,   ecology  and 

and  behavior  of  spiny  lobsters,  Panulirus  art/us,  of  St.  John,  U. S.V.I.:    (IV)    Habita- 
tion,  patterns   of  movement  and  general   behavior.     Pages   31-45   in   S.   Earl   and    R. 

Lavenberg,  Eds.,  Results  of  the  Tcktitc  Program:    coral  reef  invertebrates  and  plants. 

Bulletin  of  the  Natural  History  Museum,  Los  Angeles  County,  California. 
JOHNSON,  B.,  AND  B.  ACHE,  1978.     Antennular  chemosensitivity  in  the  spiny  lobster,  Panulirus 

argns:  amino  acids  as  feeding  stimuli.    Mar.  Bchav.  Physiol.,  in  press. 
KACZMAREK,  L.  K.,  AND  A.  N.  DAVISON,  1972.     Uptake  and  release  of  taurine  from  rat  brain 

slices.   /.  Neurochem.,  19 :  2355-2362. 
KITTREDGE,  J.  S.,  F.  T.  TAKAHASHi,  J.  LiNDSEY,  AND  R.  LASHER,   1974.     Chemical  signals  in 

the  sea:    marine  allelochemics  and  evolution.    Fish.   Hull.  Nat.  Mar.  Fish.  Scrv.,  72: 

1-11. 
LAHDESMAKI,  P.,  AND  S.  S.  O.TA,   1973.     On  the  mechanism  of  taurine  transport  at  the  brain 

cell  membranes.   /.  Ncurochcw.,  20:  1411-1417. 
LAVERACK,   M.   S.,   1964.     The  antennular  sense  organs  of  Panulirus  argns.    Comp.   Biochem. 

Physiol.,  13:  301-321. 
LENHOFF,  H.  M.,  1975.     On  the  evolution  of  receptors  associated  with  feeding.     Pages  223-236 

in  R.   Galun,   P.   Hillman,   I.   Parnes,  and  R.  Werman,  Eds.,  Sensory  physiology  and 

behavior.    Plenum  Press,  N.Y.  and  London. 
MULLONEY,  B.,  AND  A.  SELVERSTON,  1974.     The  organization  of  the  stomatogastric  ganglion  of 

the  spiny  lobster.     I.  Neurons  driving  the  lateral  teeth.  /.  Comp.  Physiol..  91  :  1-32. 
ROBBINS,   J.,    1959.      The   excitation   and   inhibition   of   crustacean    muscle   by   amino   acids.    /. 

Physiol.,  148 :  39-50. 
SHEPHEARD,   P.,   1974.     Chemoreception  in  the  antennule  of  the  lobster,  Homams  ainerieanns. 

Mar.  Bchav.  Physiol.,  2  :  261-273. 
STURMAN,  J.  A.,  G.  W.  HEFNER,  A.  F.  HOFMANN,  AND  P.  J.  THOMAS,   1976.     Taurine  pool 

sizes  in  man  :    studies  with  ^S-taurine.     Pages  21-35'  in  R.  Huxtable  and  A.  Barbeau, 

Eds.,  Taurine.  Raven  Press,  New  York. 
WELTY,  J.  D.,  W.  O.  READ,  AND  K.  H.  BYINGTON,  1976.     Comparison  of  amino-sulfonic  acids 

as  antiarrythmic  agents  in  dogs.     Pages  169-173  in  R.  Huxtable  and  A.  Barbeau,  Eds., 

Taurine.  Raven  Press,  New  York. 


Reference:   Biot.  Bull..  154  :   _'41-_'M.     (  April,  1978) 


LARVAL  DE\TELPOMEXT  OF  THE  RARE  BURROWING  MUD  SHRIMP 

NAUSHONIA  C  R  AN  GO  N  01  DBS  KIXGSLEY  (DECAPODA: 

THALASSINIDEA ;  LAOMEDIIDAE) 

JOSEPH  W.  GOY  i  AND  ANTHONY  J.  PROVENZANO,  JR. 

Institute  of  Oceanography,  Old  Dominion  University,  Xorjolk,  Virginia  23508 

Thompson  (1903)  discovered  unusual  larvae  in  the  plankton  of  Woods  Hole, 
and  after  comparing  postlarvae  obtained  from  planktonic  late  stage  larvae  with  adult 
specimens  of  Naushonia  crangonoides,  he  was  able  to  attribute  the  planktonic  larvae 
to  this  species.  Adults  of  Naushonia  crangonoides  are  known  only  from  the  Woods 
Hole  region  (Williams,  1974),  but  larvae  similar  to  Thompson's  have  been  taken  in 
Delaware  Bay  (Deevey,  1960),  Narragansett  Bay  (Hillman,  1964),  and  Chesa- 
peake Bay  (Sandifer,  1972;  Goy,  1976).  Larvae  from  Chesapeake  Bay  show  a 
number  of  differences  from  the  description  given  by  Thompson.  Moreover,  at  least 
one  other  species  of  the  genus  is  known  from  western  Atlantic  waters  (Rathbun, 
1901  ;  Gurney  and  Labour,  1939). 

The  account  of  the  early  development  of  A\  crangonoides  as  given  by  Thomp- 
son is  incomplete  in  description  and  in  illustrations,  making  identifications  of  plank- 
tonic larvae  difficult.  The  purpose  of  the  present  study  is  to  provide  a  redescrip- 
tion  of  the  larval  development  of  XansJionni  crangonoides,  and  to  review  larval 
characters  in  the  family  Laomediidae. 

MATERIALS  AND  MKTIIODS 

First  stage  larvae  of  N.  crangonoides  were  taken  in  plankton  collections  off  Cape 
Henry,  Virginia,  U.S.A.  (36°  56'  45";  76°  00'  W)  on  June  30,  1976  and  July  28, 
1976.  When  brought  to  the  laboratory,  the  larvae  were  immediately  placed  in- 
dividually into  compartmented  plastic  trays  with  25  ml  of  25C/CC  artificial  sea  water 
(Instant  Ocean,  Aquarium  Systems,  Inc.,  Eastlake,  Ohio,  U.S.A.).  Freshly 
hatched  Artemia  salina  nauplii  (San  Francisco  strain)  were  added  to  the  compart- 
ments daily  as  food.  All  zoeae  were  placed  in  a  darkened  incubator  at  25°  C, 
receiving  light  only  1  to  2  hr/day  when  the  water  was  changed  and  new  food  pro- 
vided. 

Postlarvae  and  juveniles,  as  they  reached  these  stages,  were  transferred  to  8.0 
cm  diameter  culture  dishes  ( 125  ml  capacity)  containing  75  ml  of  25%o  sea  water 
and  substrate  in  the  form  of  fine  sand  or  fine  mud,  to  facilitate  observation  of  bur- 
rowing behavior. 

A  daily  record  of  molting,  mortality,  and  sequence  of  larval  stages  was  kept. 
Larvae  and  exuviae  of  known  history  were  preserved  in  7Q%  ethyl  alcohol.  To 
obtain  more  material  for  comparisons  of  appendages,  dead  animals  were  slowly 
heated  in  5%  KOH  for  approximately  ten  minutes  to  remove  tissue  from  the  exo- 

1  Present  address :  Institute  of  Marine  Science,  University  of  North  Carolina,  Morehead 
City,  North  Carolina  28557. 

241 


242 


.!.  \V.  GOY  AND  A.  J.  PROVENZANO 


skeleton.  These  specimens  and  all  casts  from  molted  animals  were  stained  in  either 
Mallory's  Acid  Fuchsin  Red  or  Chlora/ol  Black  K  (\%  in  707^  alcohol).  The 
dissection  of  the  appendages  was  done  in  lactic  acid,  followed  by  mounting  in  gly- 
cerin jelly.  Drawings  were  made  with  the  aid  of  a  camera  lucida ;  measurements 
were  made  with  the  aid  of  a  stage  micrometer.  Total  length  of  larvae,  postlarvae, 
and  jnvenile  stages  was  measured  from  the  tip  of  the  rostrum  to  the  most  posterior 
margin  of  the  telson,  and  excluded  all  telson  processes  and  setae.  Length  of 
carapace  was  measured  from  tip  of  the  rostrum  to  the  posterolateral  margin  of  the 
carapace. 

Duration  refers  to  the  time  spent  in  a  given  stage  by  zoeae  that  survived  the  molt 
to  the  succeeding  stage.  The  term  staye  is  used  here  to  refer  to  the  intermolt 
phase  of  larval  development. 

RESULTS 

First  stage  specimens  of  Nanshonia  crangonoides  reared  under  laboratory 
conditions  reached  the  postlarval  stage  after  six  or  seven  zoeal  intermolts.  Length 
measurements  for  each  prejuvenile  stage,  range  and  mean  duration  data  for  these 
stages  are  given  in  Table  I. 

Of  18  first  stage  larvae  collected  from  plankton,  four  died  without  molting  and 
four  of  the  remainder  reached  postlarva.  A  third  and  a  fourth  stage  larva  were 
also  collected  from  plankton  and  produced  postlarvae  after  molting  several  times. 
Of  the  six  postlarvae  obtained  in  the  laboratory,  only  two  molted  to  the  next  stage, 
one  of  them  reaching  fifth  juvenile  instar  in  53  days  after  metamorphosis. 

The  intermolt  period  is  approximately  seven  days  for  the  first  two  molts  after 
postlarva  and  approximately  14  days  for  each  succeeding  molt.  The  animal  in- 
creases in  total  length  by  less  than  0.5  mm  with  each  successive  molt. 

Postlarvae  were  observed  to  feed  on  Artcinia  but  not  on  detritus  provided. 
Juveniles  following  the  postlarva  fed  only  on  detritus  and  not  on  Artcinia.  No 
burrowing  behavior  was  observed. 


TABLE  I 

Range  and  mean  duration,  total  length  and  carapace  length  of  the  larval  stage  of  Naushonia  crangon- 
oides. 


Number  of 

Range  of 

Mean 

Number  of 

Total 

Carapace 

Stage 

specimens 
molting  to 
next  stage 

duration 
(days) 

duration 
(days) 

specimens 
measured 

length 
(mm) 

length 
(mm) 

I 

13 

5-8* 

5.24* 

3 

2.3-  2.6 

0.8-0.9 

II 

11 

3-5 

3.73 

2 

2.9-  3.4 

0.9-1.2 

III 

11 

3-7 

3.82 

2 

4.5-  5.0 

1.5-1.8 

IV 

13** 

3-6 

4.76 

2 

6.3-  6.6 

1.8 

V 

11 

3-9 

4.73 

7 

7.1-  8.0 

2.1-2.5 

VI 

7 

3-12 

9.28 

2 

7.8-  8.5 

2.2-2.7 

VII 

2 

6-7 

6.50 

2 

9.6-10.4 

2.8-3.7 

PL 

2 

3-4 

3.50 

6 

4.5-  4.8 

2.7-2.9 

*First  stage  zoeae  were  collected  in  the  plankton  so  this  duration  is  probably  an  underestimate. 
**  Includes  two  planktonic  specimens. 


LARVAL  DEVELOPMENT  OF  NAUSHONIA 


243 


FIGURE  1.    Nanshonia  crangonoides:   lateral  view  of  zoeal  stages  I  (A),  II  (B),  III  (C), 
and  VI  (D)  ;  dorsal  view  of  zoeal  stages  I  (E),  II  (F),  III  (G),  and  VI  (H). 


244 


J.  W.  GOY  AND  A.  J.  PROVENZANO 


FIGURE  2.     Naushonia  crain/nnoides:    telson  of  zoeal  stages  I   (A),  II   (B),  III   (C),  and 
VI  (D)  ;  antennule  of  zoeal  stages  I  (E),  II  (F),  III  (G),  and  VI  (H), 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  245 

First  zoca  (Figs.  1A,  E) 

Rostrum  small,  slender,  upturned  at  end.  Carapace  smooth  with  no  spines 
forming  short  "neck"  forward  of  mandibles.  Eyes  sessile.  Abdomen  without 
spines  but  somites  modified  with  small  procurved  pleural  hooks.  Hooks  only  im- 
perfectly developed  on  first  somite  in  earlier  stages ;  completely  absent  on  sixth 
somite  in  all  stages.  Sixth  somite  fused  to  telson.  Telson  (Fig.  2A)  triangular, 
with  deeply  notched  posterior  margin,  bearing  five  pairs  articulated  plumose  setae; 
pair  of  external  spines ;  pair  of  fine  hairs  representing  reduced  second  telson  process. 

Antennule  (Fig.  2E)  uniramous  extending  beyond  rostrum,  slightly  longer 
than  antenna ;  bearing  four  large  aesthetascs,  one  small  plumose  seta  terminally,  and 
a  long  plumose  seta  subterminally. 

Antennal  endopodite  articulated  to  basipodite  (Fig.  3A)  bearing  three  terminal 
plumose  setae.  Antennal  scale  narrow,  oval,  bearing  ten  plumose  setae  on  medial 
margin,  most  distal  being  smallest.  Basipodite  with  spine  at  base  of  scale. 

Mandibles  (Figs.  3E,  3F)  asymmetrical,  left  one  sickle-shaped,  bearing  on  inner 
surface  of  base  four  stout  teeth;  inner  apex  an  erect,  serrate  plate  bearing  seven 
teeth.  Mandible  on  right  side  conical  with  inner  surface  of  base  bearing  stout 
process  having  four  teeth  and  serrate  plate.  Paragnath  (Fig.  3e)  on  left  side  trans- 
formed into  slender  sickle ;  remaining  in  this  form  throughout  larval  development. 

Maxillule  (Fig.  4A)  with  unsegmented,  unarticulated  endopodite  with  three 
terminal  setae.  Coxal  endite  bearing  two  stout  spines  and  two  setae  terminally  and 
one  seta  subterminally,  while  basal  endite  bearing  three  setae  terminally  and  one 
subterminally. 

Maxilla  (Fig.  4E)  with  three  inner  lobes;  proximal  lobe  of  coxal  endite  ap- 
parently absent,  but  distal  lobe  present,  bearing  one  seta.  Four  setae  on  proximal 
and  distal  lobes  of  basal  endite.  Endopodite  reduced,  bearing  four  terminal  setae. 
Scaphognathite  small,  without  proximal  extension,  bearing  five  short  plumose  setae. 

First  maxilliped  (Fig.  41)  with  four  setae  along  medial  margin  of  basipodite. 
Exopodite  bearing  four  long  plumose  natatory  setae.  Endopodite  with  four  seg- 
ments ;  setation  proximal  to  distal,  2-1-2-4.  This  formula  unchanged  in  later 
stages. 

Second  maxilliped  (Fig.  5A)  with  exopodite  bearing  four  plumose  natatory 
setae.  Endopodite  four-segmented,  setation  proximal  to  distal  0-0-2-4.  This 
formula  unchanged  in  later  stages. 

Third  maxilliped  (Fig.  5E)  two-jointed  rudiment  without  setae. 

The  chromatophores  are  mostly  red,  small,  numerous,  and  dispersed  over  the 
whole  animal.  Smaller  yellowish  chromatophores  blend  with  the  red  ones  to  pro- 
duce an  overall  orange  or  ruddy  tint.  Concentrations  of  red  pigment  are  found 
on  the  antennule  and  the  ventral  surface  of  the  carapace,  abdomen,  and  telson. 
There  is  a  yellow  cast  on  the  antennae,  mandibles,  maxillae,  the  medial  surface  of 
the  maxillipeds,  and  the  dorsal  surfaces  of  the  carapace,  abdominal  somites,  and 
telson.  This  pattern  is  also  typical  of  later  larval  stages.  By  the  postlarval  stage, 
the  animal  is  almost  colorless  with  the  ruddy  background  color  largely  lost,  and 
the  red  chromatophores  very  contracted. 


246 


J.  W.  GOY  AND  A.  J.  PROVENZANO 


FIGURE  3.  Naushonia  crangonoides:  antenna  of  zoeal  stages  I  (A),  II  (B),  III  (C),  and 
VI  (D)  ;  mandibles  of  zoeal  stage  I,  left  (E),  right  (F),  and  left  paragnath  (e)  ;  left  and 
right  mandibles  of  zoeal  stages  II  (G,  H),  III  (I,  J),  and  VI  (K,  L). 


LARVAL  DEVELOPMENT  OF  NAUSHONIA 


247 


FIGURED  Naushonia  crangonoidcs:  maxillule  of  zoeal  stages  I  (A),  II  (B),  III  (C),  and 
VI  (D)  ;  maxilla  of  zoeal  stages  I  (E),  II  (F),  III  (G),  and  VI  (H)  ;  first  maxilliped  of 
zoeal  stages  I  (I),  II  (J),  III  (K),  and  VI  (L). 


248 


J.   \V.  GOY  AND  A.  J.  PROVENZANO 


FIGURE  5.    Naushonia  crangonoides:  second  maxilliped  of  zoeal  stages  I   (A),  II   (B),  III 
(C),  and  VI  (D)  ;  third  maxilliped  of  zoeal  stages  I  (E),  II  (F),  III  (G),  and  VI  (H). 


Second  soca  (Figs.  IB,  F~) 

General  form  of  carapace  and  abdomen  as  in  first  zoea,  except  eyes  mobile. 
Telson  notch  (Fig.  2B)  less  pronounced,  with  one  minute  lateral  spine  on  each 
external  spinous  process.  Spine  2  reduced  to  thalassinid  hair  and  rather  indistinct. 
Either  five  or  six  pairs  of  plumose  setae  on  posterior  margin  of  telson. 

Antennular  peduncle    (Fig.   2F)    three-segmented  and   biramous,   with   bud   of 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  249 

inner  flagellum  carrying  a  small  })lumose  seta;  hud  of  outer  flagellum  bearing  four 
or  five  aesthetascs  and  fine  setule.  A  long  plumose  seta  located  medially  on  distal 
segment. 

Armature  of  antennal  endopodite  (Fig.  3B)  a  single  minute  fine  setule.  An- 
tennal  scale  with  11  or  12  plumose  setae  medially;  two  spines  on  hasipodite  at  hase 
of  scale  and  endopodite. 

Ten  teeth  on  serrate  plate  of  mandihles  (Figs.  3G,  3H)  ;  stout  spinous  teeth 
reduced  to  two. 

Maxillule  (Fig.  4B)  unchanged  except  endopodite  now  articulated. 

Maxilla  (Fig.  4F)  with  three  setae  on  coxal  endite,  three  and  four  on  proximal 
and  distal  basal  endites,  respectively.  Endopodite  with  one  seta ;  scaphognathite 
with  ten  plumose  setae  on  margin. 

First  and  second  maxillipeds  (Figs.  4],  5B)  unchanged  in  setation  but  slightly 
larger. 

Third  maxilliped  (Fig.  5F)  consisting  of  exopodite  bearing  four  plumose  setae, 
and  rudimentary  endopodite  subterminally  on  basipodite. 

Chelipeds  (first  pereiopods)  uniramous.  two-jointed  rudiments  without  setae. 

Second,  third,  and  fourth  pereiopods  uniramous  buds. 

Third zoea  (Figs.  1C,G} 

Sixth  abdominal  somite  distinct,  uropods  developed.  Exopodite  of  uropod  bear- 
ing 12  to  16  plumose  setae,  endopodite  a  bud  without  setae.  Posterior  margin  of 
telson  straighter,  no  indentation  (Fig.  2C)  ;  one  minute  lateral  spine  on  each  ex- 
ternal spinous  process,  but  thalassinid  hair  absent;  seven  or  eight  pairs  of  plumose 
processes. 

Antennular  peduncle  (Fig.  2G)  three-segmented  and  biramous,  bud  of  inner 
flagellum  slightly  longer  than  outer,  bearing  minute  setule  terminally.  Bud  of  outer 
flagellum  with  four  aesthetascs  terminally  and  a  short  plumose  seta  subterminally. 
Distal  segment  bearing  two  long  plumose  setae  on  inner  margin  and  short  plumose 
seta  terminally  on  outer  margin.  Middle  segment  bearing  three  long  plumose  setae 
on  inner  margin  and  two  sets  of  two  short  plumose  setae  on  outer  margin. 

Antenna  (Fig.  3C)  with  a  terminal  minute  setule  on  endopodite;  antennal  scale 
with  15  to  17  plumose  setae  medially.  Two  spines  on  basipodite  at  base  of  scale 
and  endopodite. 

Mandibles  (Figs.  31,  3J)  unchanged  but  larger. 

Endopodite  of  maxillule  (Fig.  4C)  still  unsegmented  and  bearing  three  to  five 
setae  terminally.  Basal  endite  bearing  three  to  four  short  setae  and  four  short 
spines ;  coxal  endite  bearing  two  to  four  setae  terminally,  one  subterminally. 

Maxilla  (Fig.  4G)  with  two  or  three  setae  on  coxal  endite,  four  to  six  setae  on 
each  lobe  of  basal  endite.  Endopodite  with  one  seta;  scaphognathite  bearing  12  to 
1 5  plumose  setae. 

First  maxilliped  (Fig.  4K)  with  three  or  four  setae  along  medial  margin  of 
basipodite,  exopodite  with  five  or  six  long  plumose  setae.  Endopodite  four-seg- 
mented ;  setation.  proximal  to  distal.  2-1-2-4. 

Second  maxilliped  (Fig.  5C)  with  setation  of  exopodite  increased  to  five  or  six. 

Third  maxilliped  (Fig.  5G)  with  five  or  six  natatory  setae. 


250 


J.  W.  GOY  AND  A.  J.  PROVENZANO 


FIGURE  6.     Naushonia  crangonoides:    postlarva,  lateral  view   (A),  dorsal  view   (B),  first 
pereiopod  (C),  and  second  pereiopud  (Dj. 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  251 

First,  second,  and  third  pereiopods  biramous  without  functional  exopodites. 
Fourth  and  fifth  pereiopods  still  unirann  >u.s  lnuls. 

Fourth  zoea 

Exopodites  on  first  three  anterior  pereiopods  now  functional.  Endopodite  of 
uropod  articulated  with  9  to  12  plumose  setae;  exopodite  bearing  15  plumose  setae. 
Telson  unchanged,  but  each  external  spinous  process  now  bearing  two  minute  lateral 
spines. 

Antennular  peduncle  three-segmented ;  bud  of  inner  flagellum  bearing  a  fine 
minute  setule  and  extending  further  than  outer  flagellum.  Bud  of  outer  flagellum 
bearing  five  aesthetascs,  three  short  plumose  setae  terminally,  and  one  long  plumose 
seta  subterminally  on  inner  margin.  Distal  segment  bearing  two  long  feathered 
setae  on  inner  margin  and  a  short  feathered  seta  terminally  on  outer  margin. 
Middle  segment  with  four  long  plumose  setae  on  inner  margin  and  two  sets  of  two 
short  plumose  setae  on  outer  margin. 

Antennal  endopodite  extending  beyond  antennal  scale  with  terminal  setule ;  scale 
bearing  16  to  18  plumose  setae. 

Mandibles  unchanged  but  larger. 

Endopodite  of  maxillule  may  have  up  to  six  setae.  Basal  endite  with  six  to  nine 
spinous  processes,  while  coxal  endite  with  three  to  six. 

Maxilla  (Fig.  7K)  bearing  three  to  five  setae  on  each  of  three  inner  lobes. 
Endopodite  bearing  two  to  three  setae;  scaphognathite  with  13  or  14  plumose  setae. 

First,  second,  and  third  maxillipeds  unchanged 

Exopodites  of  first,  second,  and  third  pereiopods  functional,  with  6—6—5  long 
plumose  setae,  respectively.  Fourth  and  fifth  pereiopods  uniramous  buds. 

Fifth  zoea 

Principal  distinguishing  feature :  all  pereiopods  except  fifth  now  bearing  func- 
tional exopodites.  Exopodite  of  uropod  with  22  to  28  plumose  setae ;  endopodite 
with  15  to  23  plumose  setae.  Telson  spinous  process  unchanged;  posterior  margin 
straighter  than  in  preceding  stage. 

Antennular  peduncle  two-segmented,  bud  of  inner  flagellum  bearing  fine  minute 
setule  and  slightly  longer  than  outer.  Bud  of  outer  flagellum  bearing  only  five 
aesthetascs.  Distal  segment  of  peduncle  bearing  four  or  five  long  plumose  setae  on 
inner  margin  and  two  small  plumose  setae  terminally  on  outer  margin.  Proximal 
segment  with  four  or  five  long  plumose  setae  on  inner  margin  and  three  sets  of 
shorter  plumose  setae  (3-2-2)  on  outer  margin. 

Antennal  endopodite  without  terminal  minute  setule.  Antennal  scale  shorter 
than  endopodite  bearing  19  or  20  plumose  setae  medially. 

Mandibles  larger,  otherwise  unchanged. 

Endopodite  of  maxillule  unchanged.  Basal  endite  with  six  to  nine  short  pro- 
cesses ;  coxal  endite  with  five  or  six  short  setae. 

Maxilla  with  three  setae  on  coxal  endite.  three  to  five  on  lobes  of  basal  endite. 
Endopodite  with  three  or  four  setae  and  scaphognathite  with  15  to  22  plumose  setae. 

First,  second,  and  third  maxillipeds  unchanged,  but  larger. 


252 


J.  W.  GOY  AND  A.  J.  PROVENZANO 


FIGURE  7.     Nanshonia  crangonoidcs:    postlarval  appendages:   telson    (A),   antennule    (B), 
antenna   (C),  antennal  scale   (c),  mandible   (D),  maxillule   (K),  and  maxilla   (F). 


LARVAL  DEVELOPMENT  OF  NAUSIIONIA  253 

Exopodites  of  first  three  pereiopods  unchanged,  but  fourth  pereiopod  biramous 
with  five  long  plumose  setae  on  exopodite;  fifth  pereiopod  unirumous  and  rudi- 
mentary. 

Pleopods  present  on  second  to  fifth  abdominal  somites  as  minute  buds,  hardly 
discernible. 

Sirth  soca  (Figs.  ID,  III) 

Most  distinguishing  feature:  pleopods  nonfunctional  and  without  setae,  but  now 
larger  and  biramous.  Five  animals  out  of  six  molted  to  postlarvae  from  this  stage. 
Exopodite  of  uropod  with  20  to  22  plumose  setae;  endopodite  with  17  to  20  plumose 
setae.  Telson  (Fig.  2D)  with  two  minute  lateral  spines  on  each  external  spinous 
process  and  seven  pairs  of  plumose  spines  on  posterior  margin. 

Antennular  peduncle  (Fig.  2H)  two-segmented,  bud  of  inner  flagellum  un- 
changed. Bud  of  outer  flagellum  bearing  two  fine  setules  in  addition  to  five 
aesthetascs.  Distal  segment  bearing  three  long  plumose  setae  on  inner  margin, 
one  long  plumose  seta  medially  and  terminally,  and  two  small  plumose  setae  termi- 
nally on  outer  margin.  Proximal  segment  with  six  long  plumose  setae  on  inner 
margin,  two  sets  of  three  short  feathered  setae  on  outer  margin,  and  four  plumose 
setae  medially,  proximal  to  base. 

Antennal  endopodite  unchanged.  Antennal  scale  with  20  plumose  setae  medially 
(Fig.  3D). 

Number  of  teeth  on  serrate  plate  of  mandibles  (Figs.  3K,  3L)  increased  from 
1C)  to  14. 

Maxillule  (Fig.  4D)  with  four  terminal  spines  and  one  subterminal  spine  on 
coxal  endite,  five  to  nine  processes  on  basal  endite,  endopodite  with  three  to  five 
setae. 

Maxilla  (Fig.  4H)  with  setation  of  lobes  on  endites  essentially  unchanged. 
Endopodite  bearing  three  or  four  setae;  scaphognathite  with  18  or  19  plumose 
setae. 

Setation  of  maxillipeds  (Figs.  4L,  5D,  5H)  unchanged. 

Pereiopods  unchanged. 

Pleopods  now  biramous  but  lacking  setae. 

Seventh  zoea 

This  stage  was  seen  in  four  animals  :  two  died  in  this  stage,  one  molted  to  post- 
larva,  and  the  fourth  molted  three  additional  times  within  30  days  but  never  reached 
postlarva. 

Exopodite  and  endopodite  of  uropod  each  bearing  25  plumose  setae.  Telson 
in  some  specimens  having  single  medial  spine  in  place  of  medial  pair. 

Antennular  peduncle  two-segmented,  bud  of  outer  flagellum  now  bearing  six 
aesthetascs.  Distal  segment  of  peduncle  now  bearing  four  long  plumose  setae  on 
inner  margin. 

Marginal  setae  of  antennal  scale  increased  to  22. 

Serrate  plate  of  mandibles  with  12  to  14  teeth. 

Endopodite  of  maxillule  unchanged.  Basal  endite  bearing  stout  spines ;  coxal 
endite  with  six  short  setae  terminallv  and  two  subterminallv. 


254 


J.  W.  GOY  AND  A.  J.   I'KOVKN/ANn 


Maxilla  with  two  to  four  terminal  setae  on  coxal  endite,  four  to  six  terminal  on 
proximal  lobe  of  basal  endite,  and  four  or  live  terminal  setae  on  distal  lobe.  Endo- 
podite bearing  three  setae.  Scaphognathite  \vith  16  to  18  plumose  setae. 

First  and  second  maxillipeds  unchanged,  but  larger. 

Endopodite  of  third  maxilliped  hearing  three  short  spines  subterminally  on  outer 
margin. 

Exopodites  of  first  four  pereiopods  bearing  six  long  plumose  setae  on  terminal 
segment.  Endopodite  of  first  pereiopod  slightly  swollen  with  slight  fissure  dis- 
cernible under  high  power  (trace  of  chela?). 

Fifth  pereiopod  uniramous,  very  elongate  and  slender,  bearing  minute  spine  on 
apex. 


1mm 


FIGURES.     Naushonia  crangonnidcs:    postlarval  appendages:    first  maxilliped   (A),  second 
maxilliped  (B),  and  third  maxilliped  (C). 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  255 

Pleopods  on  second  to  fifth  abdominal  somites,  biramous,  with  single  long  plu- 
mose setae  on  each  exopodite. 

Postlarra  (Figs.  7 A,  7B} 

Carapace  with  strong  supra-antennal  spine,  linea  thalassinica  and  cervical 
groove  strongly  marked.  Eyes  visible  from  above.  Rostrum  with  lateral  serrations. 
Telson  (Fig.  7A)  with  small  lateral  spine  and  42  plumose  setae  on  posterior  mar- 
gin. Serrate  transverse  sutures  on  endopodites  and  exopodites  of  uropods  ending 
with  external  spine.  Endopodite  bearing  three  spines  dorsally  and  44  plumose 
setae  (in  each  of  two  specimens),  while  exopodite  also  with  three  dorsal  spines  but 
48  plumose  setae. 

Peduncle  of  antennule  (Fig.  7R)  consisting  of  four  segments  extending  far 
beyond  front  of  eye.  Proximal  segment  bearing  numerous  setae  around  terminal 
edge ;  second  segment  with  six  setae  and  a  prominent  spine  on  inner  margin,  seven 
other  setae  dorsally.  Next  two  segments  with  four  setae  and  eight  setae,  re- 
spectively, on  ventral  surface.  External  flagellum  of  five  segments  each  bearing  one 
aesthetasc  and  two  setae,  except  penultimate  segment  with  two  aesthetascs  and 
three  setae.  Four-segmented  inner  flagellum  with  two  setae  on  three  proximal 
segments  and  four  setae  on  terminal  segment. 

Second  segment  of  antenna!  peduncle  bearing  ovate  antennal  scale  (Figs.  7C, 
7c)  with  six  lateral  teeth,  six  lateral  setae,  and  15  plumose  setae  on  inner  margin. 
Second,  third,  and  fourth  peduncular  segments  with  spine  distally  on  inner  margins. 
Flagellum  of  approximately  48  segments,  most  bearing  setae  distally. 

Mandibles  (Fig.  7D)  symmetrical,  cutting  edge  with  four  small  teeth.  Palp 
developed,  unsegmented,  bearing  minute  seta  terminally. 

Endopodite  of  maxillule  (Fig.  7E)  with  minute  seta  terminally,  rarely  with  seta 
proximally.  Basal  endite  rounded  and  bearing  two  long  terminal  plumose  setae 
and  14  short  spines.  Coxal  endite  with  nine  short  spines. 

Maxilla  (Fig.  7F)  showing  proximal  lobe  of  coxal  endite  for  the  first  time, 
with  three  long  and  eight  short  setae ;  distal  lobe  of  coxal  endite  and  proximal  lobe 
of  basal  endite  each  with  eight  setae,  distal  lobe  of  basal  endite  with  12  setae. 
Unsegmented  endopodite  bearing  one  long  and  four  short  setae  terminally,  two 
setae  subterminally.  Scaphognathite  with  37  plumose  setae  on  outer  margin  and 
15  minute  setae  on  inner  margin. 

First  maxilliped  (Fig.  8 A  )  with  two-lobed  basipodite,  proximal  lobe  bearing  13 
plumose  setae,  distal  lobe  bearing  12  marginal  and  12  submarginal  plumose  setae. 
Endopodite  three-segmented  with  terminal  segment  not  expanded  without  setae. 
First  segment  with  11  feathered  setae,  second  (middle)  segment  with  three 
feathered  setae  on  outer  margin.  Exopodite  three-segmented  with  widened  proxi- 
mal segment  bearing  10  plumose  outer  setae.  Second  segment  lacking  setae,  distal 
segment  with  two  long  plumose  terminal  setae.  Elongate  epipodite  bearing  long 
plumose  terminal  seta. 

Basipodite  of  second  maxilliped  (Fig.  8B)  with  three  long  plumose  setae  on 
inner  margin,  heavily  serrate  epipodite  on  outer  margin.  Endopodite  four-seg- 
mented with  4—0-4—10  setae,  proximally  to  distally.  Exopodite  two-segmented 


256 


J.   \V.  GOY  AND  A.  J.  PROVENZANO 


TABLE  II 
Comparison  of  some  postlarval  characters  of  Naushonia  crangonoides  and  Naushonia  portoricensis. 


N.  crangonoides 


N.  portoricensis 


Rostrum 

Rostral  apical  process 

Linea  thalassinica 

Small,  anterior  tubercle  on  eye 

Ischium  of  mxps 

Arthrobranch  mxpi 

First  pereiopod 

Tooth  of  propodus  of  pi 

Dactyl  of  pi 

Marginal  setae  of  telson 

Suture  of  uropodal  endopodite 


No  lateral  teeth 

Absent 

Distinct 

Absent 

Serrate 

Present 

Does  exceed  eyes 

Two  large,  four  small 

\  length 

42 

Complete 


Lateral  tooth  on  each  side 

Present 

Faint 

Present 

Smooth 

Absent 

Does  not  exceed  eyes 

One  large,  inner 

\  length 

22 

Incomplete 


with  eight  short  setae  on  first  segment  and  four  long  plumose  terminal  setae  on 
second  segment. 

Basipodite  of  third  maxilliped  (Fig.  <SC)  with  numerous  short  plumose  setae, 
one  long  plumose  seta.  Epipodite  consisting  of  serrate-margined  mastigobranch 
with  two  plumose  setae,  rudimentary  podobranch.  Endopodite  five-segmented, 
ischium  serrate  distally,  hearing  four  short  plumose  setae.  Number  of  setae  on 
other  segments  as  follows:  four  on  inner  margin  and  two  on  outer  margin  of 
second  segment;  three  on  inner  margin  and  three  terminally  on  third  segment;  six 
on  inner  margin,  seven  subterminally,  and  four  terminally  on  fourth  segment ;  and 
six  terminally  with  five  shorter  setae  dispersed  randomly  on  fifth  segment.  Exo- 
podite  two-segmented,  with  four  long  plumose  terminal  setae. 

Chelipeds  (Fig.  6C)  slender,  symmetrical  extending  well  beyond  cornea,  held 
stiffly  in  front  of  living  animals  during  locomotion.  Propodus  with  two  large,  four 
smaller  inner  teeth  ;  dactyl  slender  and  falcate. 

Second  to  fifth  pereiopods  slender,  short,  very  similar  in  structure,  except  for 
dactyls.  Second  and  third  pereiopods  (Fig.  6D)  with  robust  dactyls  bearing  five 
small  stout  spines  on  their  inner  margins,  dactyls  of  fourth  and  fifth  pereiopods 
elongate,  slender,  and  lacking  stout  spines. 

Pleopods  on  abdominal  somites  two  to  five,  biramous,  lanceolate.  Second  and 
third  pleopods  with  eleven  and  thirteen  plumose  setae  on  endopodites  and  exopo- 
dites,  respectively;  fourth  pleopod  with  nine  and  ten  setae,  fifth  pleopod  with  seven 
and  nine  setae  on  endopodites  and  exopodites,  respectively.  No  appendix  interna 
or  appendix  masculina  present. 


Juvenile  stages 

Tn  the  first  few  molts  after  the  postlarval  stage,  neither  length  of  the  specimens 
nor  their  morphology  changes  drastically.  There  seems  to  be  a  gradual  develop- 
ment of  adult  characteristics.  A  detailed  description  of  the  fifth  stage  juvenile  and 
its  appendages  and  comparison  of  this  juvenile  with  adult  Naushonia  will  be  pre- 
sented elsewhere  (Goy  and  Provenzano,  in  preparation). 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  257 

DISCUSSION 

Since  N.  crangonoides  can  reach  postlarva  in  six  zoeal  stages,  the  seventh  stage 
seems  not  to  be  essential  in  the  larval  life  history  of  Naushonia.  Thompson  found 
the  equivalent  of  the  seventh  stage  in  the  plankton,  and  we  obtained  it  four  times 
in  the  laboratory.  One  of  the  seventh  stage  zoeae  in  the  present  study  molted  an 
additional  three  times  within  30  days,  but  failed  to  metamorphose.  Studies  on 
other  larvae  show  that  in  more  than  15  families  of  decapods,  the  number  of  larval 
instars  preceding  metamorphosis  is  variable  (see  Knowlton,  1974,  for  review). 

The  pattern  of  change  in  mean  duration  as  development  proceeds  in  this  species 
(Table  I)  is  similar  to  that  observed  in  some  other  decapods.  The  first  stage  is 
longer  than  the  succeeding  several  stages,  but  towards  the  end  of  development  dura- 
tion increases  again  (for  example,  Provenzano,  1968;  Robertson,  1968).  The 
extent  of  distribution  of  this  pattern  is  not  yet  established,  but  it  probably  occurs 
only  in  species  with  moderately  long  larval  development,  and  possibly  only  in 
laboratory-reared  larvae.  Relatively  long  duration  of  the  first  stage  may  be  related 
to  the  feeding  behavior  or  yolk  supply  of  this  stage,  while  the  lengthening  of  later 
stages  over  intermediate  stages  may  be  related  to  the  energy  demands  for  tissue 
growth  and  preparation  for  metamorphosis. 

Adults  of  Naushonia  crangonoides  are  known  only  from  Massachusetts  at  Bass 
River,  Vineyard  Sound,  and  Elizabeth  Islands  ("Williams,  1974).  Larvae  believed 
to  belong  to  N.  crangonoides  have  been  collected  from  the  Woods  Hole  area  during 
July,  August,  and  September  (Fish,  1925)  ;  in  Delaware  Bay  from  August  to 
October  (Deevey,  1960)  ;  in  Narragansett  Bay  in  August  (Hillman,  1964)  ;  and  in 
Chesapeake  Bay  from  August  to  September  (Sandifer,  1972;  Goy,  1976).  In  the 
last  two  collections,  first  stage  larvae  of  Naushonia  were  most  numerous  especially 
near  the  bay  mouth.  The  presence  of  early  larval  stages  suggests  a  breeding  popu- 
lation of  Ar.  crangonoides  somewhere  near  the  mouth  of  Chesapeake  Bay. 

The  first  stage  larva  of  Naushonia  crangonoides  has  not  yet  been  hatched  from 
ovigerous  adults.  Our  identification  of  this  larval  series  is  based  on  morphological 
similarities  between  the  fifth  juvenile  stage  and  adult  specimens  of  Naushonia 
crangonoides  and  N.  portoricensis  from  collections  in  th  U.S.  Xationeal  Museum 
(Goy  and  Provenzano,  in  preparation). 

Thompson  (1903)  made  no  mention  of  the  thalassinid  hairs  on  the  telson  in  the 
first  and  second  larval  stages  of  A',  crangonoides.  These  are  extremely  minute  and 
he  probably  overlooked  them.  He  also  apparently  overlooked  the  minute  lateral 
spines  on  the  external  spinous  process  of  the  telson  of  the  third  to  the  seventh  zoea. 
Thompson  mentioned  that  no  traces  of  the  cheliped  could  be  found  even  in  his  fifth 
stage  larva,  but  close  examination  of  our  seventh  zoea,  equivalent  to  his  fifth,  showed 
under  high  power  a  slight  fissure,  probably  a  trace  of  the  chela.  Discrepancies  in 
setation  and  number  of  spines  or  teeth  on  certain  appendages  of  Thompson's  larvae 
and  those  of  the  present  study  probably  can  lie  attributed  to  normal  variation 
within  the  species.  This  is  also  probably  true  of  the  size  differences  found.  The 
zoeae  of  Thompson  were  usually  much  larger  than  those  obtained  by  us  in  the 
laboratory.  Thompson's  second  stage  larva  had  a  total  length  of  4.0  mm,  whereas 
our  largest  second  stage  zoea  was  only  3.4  mm.  The  largest  third  stage  larva  at- 
tained was  5.0  mm,  while  Thompson's  equivalent  stage  was  over  5.0  mm  and  its 


258  J.  W.  GOY  AND  A.  J.  PROVENZANO 

first  pereiopod  had  an  exopodite  functional  as  a  swimming  organ.  The  fourth  and 
sixth  stages  we  obtained  in  the  laboratory  were  omitted  from  Thompson's  descrip- 
tion entirely.  His  fourth  stage  is  equivalent  to  the  fifth  stage  in  the  present  study, 
and  his  fifth  stage  is  essentially  the  same  as  the  seventh  zoea  that  we  obtained  in 
culture. 

Larval  stages  related  to  or  belonging  to  Nanslionia  have  been  found  off  Samoa 
and  the  Great  Barrier  Reef  (Gurney,  1938)  ;  off  Bermuda  (Gurney  and  Lebour, 
1939);  and  off  New  South  Wales  (Dakin  and  Colefax,  1940).  the  larvae  re- 
ported from  Samoa  and  the  Great  Barrier  Reef  by  Gurney  have  the  essential  char- 
acters of  Naushonla,  but  are  quite  distinct  from  Ar.  crangonoidcs.  Both  Gurney 
(1938)  and  Dakin  and  Colefax  (1940)  suggested  that  Gurney's  larvae  may  repre- 
sent a  genus  other  than  Nauslwnia.  The  first  stage  larva  from  Samoa  agrees  fairly 
closely  with  Ar.  crangonoides  but  differs  in  the  absence  of  pleural  spines  on  the 
abdomen  and  the  possession  of  a  papilliform  process  on  the  fifth  somite.  The  Bar- 
rier Reef  specimen  was  believed  to  be  a  fourth  stage  larva  of  the  Samoan  species ; 
it  also  lacks  the  abdominal  pleural  spines  and  has  no  exopodite  on  the  fourth 
pereiopod.  Gurney  and  Lebour  (1939)  stated  that  if  these  same  larvae  belong 
to  Nauslwnia,  they  could  possibly  be  zoeae  of  N .  pcrrciri  known  from  the  Red  Sea. 

The  zoeae  that  bear  the  closest  resemblance  to  those  of  N.  crangonoidcs  were 
collected  from  Newr  South  Wales  by  Dakin  and  Colefax  (1940).  They  described 
six  zoeal  stages  in  their  species  of  NaiisJionia  and  recorded  a  sixth  stage  larva  just 
about  to  metamorphose,  in  which  they  could  observe  the  postlarval  telson.  The 
differences  between  that  species  and  7V.  crangonoidcs  are  slight ;  the  telson  is  shaped 
differently,  and  the  endopodite  of  the  third  maxilliped  in  later  stages  bears  a  long 
seta,  missing  in  N.  crangonoidcs. 

The  developing  telson  of  the  postlarvae  observed  by  Dakin  and  Colefax  differs 
from  that  of  N.  crangonoides  in  having  only  20  setae  on  its  posterior  margin  (42  in 
N.  crangonoidcs}  and  no  evidence  of  an  external  spine. 

The  larvae  collected  off  Bermuda  were  believed  by  Gurney  and  Lebour  (1939) 
to  belong  to  Naushonia  portoriccnsis  based  on  similarities  between  the  postlarva 
and  adults  of  that  species.  These  larvae  closely  resemble  those  of  N.  crangonoidcs 
but  differ  from  N.  crangonoidcs  in  the  relatively  smaller  size  of  N.  portoriccnsis, 
differently  shaped  telson,  and  a  stronger  pleural  spine  on  the  first  abdominal  somite. 
Gurney  and  Lebour  also  described  a  postlarval  stage  and  first  juvenile  stage  for 
what  they  believed  to  be  N.  portoriccnsis.  Their  postlarva  differs  from  ours  (Table 
II). 

Larvae  of  the  other  two  genera  of  the  Family  Laomediidae  are  known.  Cer- 
tainly identified  larva  of  Ja.rca  have  been  described  for  two  species,  /.  nocturna 
(Clans,  1884;  Cano.  1891;  Bouvier,  1914;  Caroli,  1924),  and  /.  novaesealandiae 
(Gurney,  1924;  Wear  and  Yaldwyn,  1966).  The  long-necked  Lucifcr-Yike  larva  of 
/.  nocturna  was  given  the  name  trachelifer  by  Brooks  (1889).  The  postlarva  of 
this  species  was  obtained  in  the  laboratory  from  metamorphosed  last  stage  zoeae 
(Caroli,  1924;  Tattersal,  1938).  Wear  and  Yaldwyn  (1966)  described  the  com- 
plete larval  life  history  of  Ja.rca  novaezealandi.de  from  plankton  material  and  de- 
vised keys  to  separate  larval  species  of  Ja.vca  and  first  postlarval  stages  of  /. 
novaezcalandiac,  J.  nocturna,  Nauslwnia  crangonoidcs,  and  Ar.  portoriccnsis. 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  259 

A  sixth  stage  larva  collected  by  Kurian  (1956)  in  the  Adriatic  Sea  and  at- 
tributed to  Ja.i'ca  seems  to  bear  a  closer  resemblance  to  larvae  of  Naushonia  than 
to  those  of  Ja.vca.  The  rostrum  of  Kurian's  larva  has  a  "double  curve,"  and  the 
end  of  the  rostrum  does  not  reach  the  extremity  of  the  eye,  characteristics  of  all 
known  Naushonia  larvae.  The  telson  has  12  posterior  marginal  plumose  setae 
with  two  small  spines  on  the  outer  margins  of  the  lateral  prolongations  of  the  tel- 
son;  the  telson  also  has  the  posterior  corners  drawn  into  curved  processes.  This 
is  similar  to  the  telson  of  the  XausJionia  sp.  found  by  Dakin  and  Colefax  (1940)  and 
that  of  AT.  portoriccnsis  (Gurney  and  Lebour,  1939).  The  pleural  hooks  are  ab- 
sent on  the  sixth  abdominal  somite  of  Kurian's  larva  as  in  all  known  Naushonia 
larvae.  They  are  present  on  the  larvae  of  Ja.vca  nocturna  and  /.  novaezealandiac. 
The  pleural  hooks  are  also  reduced  on  the  first  abdominal  somite,  which  is  char- 
acteristic of  Ar.  crangonoides  larvae.  The  second  antenna  of  this  Adriatic  species 
is  also  very  similar  to  that  of  Ar.  crangonoides  larvae.  This  larva  found  by  Kurian 
is  probably  only  a  stage  V.  Its  size  of  5.7  mm  fits  in  the  range  of  5.0  to  8.0  mm 
total  length  from  known  Naushonia  stage  Y  larvae,  whereas  it  is  rather  small 
for  that  of  the  known  Ja.vca  stage  Y  larvae,  which  range  from  10.7  to  12.5  mm  in 
total  length. 

The  larvae  of  the  genus  Laoincdia  are  known  from  the  first  stage  zoeae  hatched 
from  an  ovigerous  Laoincdia  astacina.  Sakai  and  Miyake  (1964)  described  these 
larvae  and  compared  them  with  zoeae  of  Xaiislwnia  and  Ja.vca.  In  their  paper  on 
L.  healvi,  Yaldwyn  and  Wear  (1972)  believed  larvae  described  by  Dakin  and 
Colefax  (1940)  from  Sydney  harbor  might  be  zoeae  of  Laoincdia  liealyi.  To  test 
this  hypothesis,  they  asked  Dr.  Sakai  for  larval  material  of  /..  astacina,  but  from 
the  description  and  illustrations  of  their  borrowed  material  one  can  see  they  mis- 
takingly  described  a  larva  of  Upogchia  instead  of  Laoincdia.  We  consider  the 
description  by  Sakai  and  Miyake  to  represent  the  true  first  stage  zoea  of  Laoincdia 
astacina. 

The  first  stage  larvae  of  the  three  genera  of  Laomediidae  can  be  distinguished. 
Laoincdia  differs  from  the  others  in  having  a  telson  formula  of  6  +  6  rather  than 
7  +  7,  in  lacking  the  thalassinid  hair  on  the  telson,  in  having  only  five  apical  pro- 
cesses on  the  antennule  rather  than  six,  and  in  not  having  a  conspicuous  rostrum. 
Naushonia  can  be  distinguished  from  Ja.vca  by  having  an  upturned,  short  rostrum 
rather  than  a  straight  or  long  rostrum. 

All  known  larvae  of  Laomediidae  can  be  distinguished  from  other  decapod 
larvae  by  the  procurved  pleural  hooks  on  at  least  four  abdominal  somites  and  by  the 
asymmetrical  mandibles  (left  mandible  and  paragnath  drawn  out  into  sickle-shaped 
structures). 


We  thank  Steve  Morgan,  who  provided  the  first  stage  zoeae  from  plankton 
cruises  on  the  R/V  LIN  WOOD  HOLTON.  and  Dr.  Raymond  A  Fanning  for  making  pos- 
sible examination  of  specimens  of  Naushonia  crangonoides  and  Naushonia  portori- 
censis  from  the  United  States  National  Museum.  This  work  was  supported  by 
National  Science  Foundation  grant  DEB76-11716. 


2nd  J.  W.  GOY  AND  A.  J.  PROVENZANO 

SUMMARY 

1.  Larval  stages  captured  from  plankton  in  Chesapeake  Kay  were  reared  in  the 
lahoratory  to  and  hevond  metamorphosis  and  were  determined  to  be  those  of  the 
mud  shrimp,  \aushoniti  crangonoides,  known  as  adults  only  from  the  area  of  Woods 
Hole. 

2.  The  postlarval  stage  in  A'tiiislioniti  may  he  reached  after  six  or  seven  zoeal 
stages.     Descriptions  and  illustrations  of  the  zoeal   stages  and  the  postlarva  are 
presented. 

3.  Larvae  are  compared   with   others  known   for   the   genus   and   the   Family 
Laomediidae  and  distinctive  specific  and  generic  characters  are  discussed. 

LITERATURE  CITED 

BOUVIER,   E.   L..    1914.      Observations   nouvelles   sur   les   Trachelifers,   larves   luciferiformes   de 

Ja.i-ea  nochirnti.    J.  Mar.  Bioi.  Assoc.  U.K.,  10(2):  194-206. 
BROOKS,  G.,   1889.     Notes  on  a   Lucifer-like  Decapod  larva   from   the  west  coast  of  Scotland. 

Proc.  R.  Soc.  Edinh.,  15(1887-88)  :    420-423. 
CANO,  G.,  1891.     Svilluppo  postembrionalle  della  Gehia,  A.rius,  Callianassa  e  Callia.vis;  Mor- 

fologie  del  Thalassinidi.     Boll.  Soc.  Nat.  Napoli  (1),  5:  5-30. 
CAROLI,  E.,  1924.     Svillupo  larvale  e  primo  stadio  postlarvale  della  Jaxca  nocturna  Nardo   (  = 

Callia.ris  adriatica  Heller).    Pubbl.  Staz.  Zoo/.  Napoli.  5:  153-197. 
CLAUS,  C,   1884.     Zur  Kenntniss  der  Kreislaufsorgane  der   Schizopoden  und  Decapoden.  Arb. 

Zoo/.  lust.  Univ.  ll'icn.,  5(3):  271-318. 
DAKIN,  W.  J.,  AND  A.  N.  COLEFAX,  1940.     Tlie  plankton  of  the  Australian  coastal  waters  off 

New  South  Wales.    Part  I.  PubL  Univ.  Sydney  Deft.  Zoo/.,  1:   1-215. 
DEEVEY,   G.   D..   1960.     The  zooplankton   of  the   surface  waters   of  the   Delaware   Bay   region. 

Bull,  liingham  Occanogr.  Coll.  Yale  Unit'..  17:  5-53. 
FISH,  C.  J.,  1925.    Seasonal  distribution  of  the  plankton  of  the  Woods  Hole  region.     Bull.  U.  S. 

Bur.  Fish.,  41 :  91-179. 
Gov,  J.  W.,   1976.     Seasonal  distribution  and  the  retention  of  some  decapod  crustacean  larvae 

within  the  Chesapeake  Bay,  Virginia.     Muster's  thesis,  Old  Dominion  University,  Nor- 
folk, Virginia,  334  pp. 
GURNEY,  R.,  1924.  Crustacea.    Part  IX — Decapod  Larvae.    Br.  Antarct.   (Terra  Noz>a)   E.rpcd. 

1910,  Zoo/..  8(2)  :  37-202. 
GURNEY,  R.,  1938.     Larvae  of  Decapod  Crustacea.     Part  V.  Nephropsidea  and  Thalassinidea. 

Discovery  Rcfi..  27  :  291-344. 
GURNEY,   R.,  AND  M.   V.   LEBOUR.   1939.     The   larvae  of  the  decapod  genus  Naushonia.    Ann. 

Mag.  Nat.  Hist.  Ser.  (11),  3(18)  :  609-614. 
HILLMAN,  N.  S.,  1964.     Studies  on  the  distribution  and  abundance  of  decapod  larvae  in  Nar- 

ragansett  Bay,  Rhode  Island,  with  consideration  of  morphology  and  mortality.   Master's 

thesis.  University  of  Rhode  Island,  Kingston,  Rhode  Island,  74  pp. 
KNOWLTON,   R.   E.,   1974.     Larval   developmental   processes  and  controlling  factors   in   decapod 

Crustacea,  with  emphasis  on  Caridea.     Thalassia  Ju(/osl.,  10(1/2)  :   138-158. 
KURIAN,   C.   V.,   1956.     Larvae  of  Decapod   Crustacea   from  the   Adriatic    Sea.    Acta  Adriat., 

6(3)  :  1-108. 
PROVENZANO,  A.  J.,  JR.,  1968.    The  complete  larval  development  of  the  West  Indian  hermit  crab 

Petrochirus  diot/enes    (L.)     (Decapoda,    Diogenidae)    reared   in   the   laboratory.    Bull. 

Afar.  Sci.,  18:  143-181. 
RATHBUN,  M.  J.,  1901.    The  Brachyura  and  Macrura  of  Porto  Rico.    Bull.  U.  S.  Fish.  Coniin 

for  1900,2:  1-127. 
ROBERTSON,    P.    B.,    1968.      The    complete    larval    development    of    the    sand    lobster,    Scyllarus 

ainericanus   (Smith)    (Decapoda,   Scyllaridae)    reared  in  the  laboratory,  with  notes  on 

larvae  from  the  plankton.    Bull.  Mar.  Sci..  18:  294-332. 
SAKAI,  K.,  AND  S.  MIYAKE,  1964.     Description  of  the  first  zoea  of  Laomedia  astacina  deHaan 

(Decapoda,  Crustacea).    Sci.  Bull.  Fac.  Agric.  Kyushu  Univ.,  21:  83-37. 


LARVAL  DEVELOPMENT  OF  NAUSHONIA  261 

SANDIFER,  P.  A.,  1972.     Morphology  and  ecology  of  Chesapeake  Bay  decapod  crustacean  larvae. 

Ph.D.  dissertation,  Ihiircrsity  of  Virginia,  Charlottesville,  Virginia,  532  pp. 
TATTERSALL,  W.  M.,  1938.    A  note  on  the  trachelifer  larva  of  Jaxea  nocturna  (Chiereghin)   and 

its  metamorphosis.    Ann.  Mag.  Nat.  Hist.  Scr.   (11),  1(6):  625-631, 
THOMPSON,  M.  T.,  1903.     A  rare  thalassinid  and  its  larva.    Proc.  Boston  Soc.  Nat.  Hist.,  31 : 

1-21. 
WEAR,  R.  G.,  AND  J.  C.  YALDWYN,  1966.     Studies  on  Thalassinid  Crustacea   (Decapoda,  Mac- 

rura,  Reptantia)  with  a  description  of  a  new  Jaxea  from  New  Zealand  and  an  account 

of  its  larval  development.    Zoo!.  Publ.   1'ictoria  Univ.   Wellington,  41:   1-27. 
WILLIAMS,  A.  B.,  1974.    Marine  flora  and  fauna  of  the  northeastern  United  States.    Crustacea : 

Decapoda.  Nat.  Oceanic  Atinos.  Admin.  Rcf.  Nat.  Mar.  Fish.  Scrv.  Circ.,  389:  1-50. 
YALDWYN,  J.   C.,  AND  R.  G.  WEAR,   1972.       The  eastern  Australian  burrowing  mud  shrimp 

Laomedia   healyi    (Crustacea,    Macrura,    Reptantia,    Laomediidae)    with    notes    on    the 

larvae  of  the  genus  Laomedia.  Aust.  ZooL,  17:    126-141. 


Reference:  Blol  Bull.,  154  :  262-281.     (April,  1978) 


REPRODUCTION  IN  THREE  SPECIES  OF  INTERTIDAL 
BARNACLES  FROM  CENTRAL  CALIFORNIA 

ANSON  H.  HINES1 
Department  of  Zoology,  University  of  California,  Berkeley,  California,  U.S.A.  94720 

The  reproductive  biology  of  eastern  Atlantic  barnacles  has  been  studied  ex- 
tensively, with  emphasis  on  the  role  of  temperature  in  regulating  reproduction  (e.g., 
Crisp,  1950,  1954;  Patel  and  Crisp,  1960a;  Barnes,  1963;  Barnes  and  Stone,  1973). 
However,  there  has  been  much  less  work  on  cirripedes  elsewhere  in  the  world. 
This  paper  compares  the  reproductive  cycles  and  brood  production  of  three  species 
of  intertidal  barnacles  abundant  in  central  California:  Chthamalns  fissiis  Darwin, 
1854;  Balanus  glandula  Darwin,  1854;  and  Tctraclita  squainosa  nibcscens  Darwin, 
1854  (hereafter  called  T.  squainosa  in  this  paper). 

On  the  west  coast  of  North  America  the  role  of  barnacles  in  the  structure  of 
intertidal  communities  has  been  stressed  (e.g.,  Connell,  1970;  Dayton,  1971),  but 
the  reproductive  cycles  of  only  two  intertidal  and  one  subtidal  species  have  received 
attention.  Balanus  glandula  from  British  Columbia  and  southern  California  broods 
primarily  in  the  cold  winter  and  spring  months  but  may  show  minor  brooding 
activity  in  summer  (Barnes  and  Barnes,  1956).  On  the  other  hand,  Pollicipes 
polymerus  has  a  variable  reproductive  cycle  with  a  limited  summer  brooding  season 
in  Washington  which  increases  in  length  to  central  California  (Hilgard,  1960; 
Lewis,  1975),  while  brooding  activity  in  southern  California  peaks  in  winter  with 
30%  still  brooding  in  summer  (Straughan,  1971).  Balanus  pacificus,  a  subtidal 
species  in  southern  California,  broods  at  high  frequencies  year-round  showing  no 
correlation  with  temperature  (Hurley,  1973).  Thus,  in  these  species  the  relation- 
ship of  reproduction  with  temperature  is  complex.  The  role  of  temperature  or 
other  environmental  factors  as  proximal  cues  synchronizing  brooding  has  not  been 
studied  experimentally  in  species  of  cirripedes  on  the  west  coast  of  North  America. 

In  this  paper  brooding  and  nutrient  storage  cycles  are  compared  in  populations 
of  the  three  species  of  barnacles  occurring  in  the  warm-water  discharge  canal  of  a 
large  power  plant  and  in  adjacent  areas  of  ambient  temperature.  Aspects  of  the 
regulation  of  these  cycles  by  temperature,  photoperiod,  and  food  availability  are 
investigated  experimentally.  The  size  and  number  of  broods  produced  during  a 
season  are  estimated  so  that  the  patterns  of  reproductive  effort  of  these  three  species 
can  be  compared  with  other  cirripedes. 

MATERIALS  AND  METHODS 

This  study  was  conducted  at  the  Pacific  Gas  and  Electric  Company  fossil-fuel 
power  plant  at  Morro  Bay,  California  (35°  22'  30"  N,  120°  52'  30"  W).  This 
1030  mega  Watt  plant  uses  ocean  water  for  once-through  cooling,  discharging  a 

1  Present  address:  Center  for  Coastal  Marine  Studies,  University  of  California,  Santa 
Cruz,  California,  U.S.A.  95064. 

262 


REPRODUCTION  IX   BARNACLES  263 

plume  with  an  isotherm  5°C  above  amlm-nt  of  about  ().(>  to  3.0  acres  surface  area. 
Continuous  temperature  records  (see  Fig.  1)  were  taken  from  "Ryan"  temperature 
recorders  positioned  next  to  the  intake  screens  and  the  outfall  next  to  the  dis- 
charge tubes.  Both  recorders  were  at  about  mean  lower  low  water,  corresponding 
to  the  approximate  intertidal  level  of  the  sampled  barnacles. 

Field  data  were  gathered  from  populations  of  Clitliunialus  fissits,  Balanits  gland- 
ula,  and  Tetraclita  sqiianiosa  occurring  in  the  warm-water  outfall  and  in  adjacent 
control  areas  of  ambient  temperatures  in  the  Morro  Bay  harbor  channel.  Col- 
lecting trips  were  made  at  about  monthly  intervals  from  November,  1972,  to 
January,  1975.  C.  fissus  is  small  (<  8  mm  basal  diameter)  and  common  in  the 
high  intertidal  Zone  1  of  Ricketts  and  Calvin  (1968)  from  San  Francisco  to  Baja 
California.  B.  glandula  grows  to  about  20  mm  in  diameter  and  is  common  in  the 
upper  mid-intertidal  Zone  2  from  Alaska  to  Baja  California.  T.  sqiianiosa  attains 
a  maximum  diameter  of  50-60  mm  and  is  found  in  lower  mid-intertidal  Zone  3 
from  San  Francisco  to  Baja  California.  Although  these  zonal  distributions  are 
characteristic,  the  three  species  are  often  found  together  in  Zone  3,  and  every  effort 
was  made  to  collect  samples  from  equivalent  tidal  levels  (0  to  +1.0  feet  above 
mean  lower  low  water)  to  minimize  effects  of  differences  in  exposure  and  feeding 
time. 

For  each  species,  barnacles  of  haphazard  sizes  over  the  entire  range  available 
were  selected  for  processing.  Each  barnacle  was  examined  for  brooded  embryos, 
ripeness  of  ovary,  and  ripeness  of  the  male  reproductive  system.  Brooding  fre- 
quencies were  calculated  for  barnacles  known  to  be  reproductively  mature.  Ovaries 
were  staged  "ripe"  when  they  had  large  quantities  of  yolky  material  bulging  into 
the  mantle  chamber,  or  "not  ripe"  when  little  or  no  yolk  was  present.  The  male 
reproductive  system  was  staged  as  "ripe"  or  "not  ripe"  according  to  the  presence  or 
absence  of  seminal  vesicles  discernibly  filled  with  white  seminal  fluid.  The  basal 
diameter  along  the  rostral-carinal  axis  and  the  following  dry  weights  were  deter- 
mined for  each  barnacle:  the  opercular  valves;  the  body  (soma  only,  excluding 
ovary,  retractor  muscles,  and  tissue  lining  the  mantle  cavity)  ;  brooded  egg  mass; 
and,  in  some  cases,  the  ovary  (including  retractor  muscles  and  tissue  lining  the 
mantle  cavity). 

Egg  numbers  per  brood  were  counted  using  a  Model  A  Coulter  Counter  modi- 
fied to  count  all  the  eggs  in  each  brood.  The  eggs  of  each  brood  were  dissociated 
with  protease  in  sea  water,  fixed  in  formalin,  and  run  through  the  counter.  The 
length  and  width  of  a  few  eggs  in  each  brood  were  recorded,  as  was  the  dry  body 
weight  of  the  parent. 

For  laboratory  experiments  barnacles  were  maintained  on  small  rocks  con- 
tinuously submerged  in  vigorously  aerated  sea  water  under  constant  photoperiod 
and  temperature  conditions.  Barnacles  were  fed  ad  lib  with  dense  suspensions  of 
Artcjnia  salina  nauplii,  augmented  in  some  cases  with  cultures  of  Dunaliella  sp. 
All  three  species  were  maintained  manv  months  using  these  techniques.  Tempera- 
ture experiments  were  conducted  at  11.5°  or  20°  C,  representing  winter  ambient 
and  outfall  water  temperatures,  respectively.  Photoperiods  used  represented  the 
long  (14L:10D),  intermediate  (12L:12D),  and  short  (10L:  141))  day-lengths 
occurring  at  Morro  Bay. 


264 


ANSON  H.  HINES 


30 


25 


U 

o 

cu 


20 


<D 
Q. 


!o 


Intake 


A        S        0        N 

1973 

D 

J 

F       M 

A 

M 

J         J 

1974 

A 

S       0 

N 

D 

J      F 
1  1975 

FIGURE  1.  Temperature  records  for  outfall  and  intake  at  Morro  Bay  power  plant,  plotting 
weekly  high,  mean,  and  low  temperatures  averaged  from  six-hour  intervals.  Recorders  were 
positioned  at  about  mean  lower  low  water. 

To  estimate  the  length  of  time  embryos  were  brooded,  egg  lamellae  judged 
freshly  deposited  were  removed  from  barnacles,  broken  into  small  chimps  of  em- 
bryos, and  held  in  the  laboratory  under  constant  temperature  and  photoperiod 
until  they  hatched  or  reached  late  developmental  stages  judged  ready  to  hatch. 
Sterilized  sea  water  for  these  in  vitro  brooding-time  experiments  was  treated  with 
antibiotics,  continuously  aerated,  and  frequently  changed. 


RESULTS 


Brooding  cycles 


Brooding  frequencies  for  outfall  and  control  populations  of  the  three  species 
are  shown  in  Figure  2.  The  control  populations  of  Clitliainaliis  fissus  brooded 
during  a  long  summer  season  from  about  March  or  April  to  October.  In  peak 
periods  from  June  to  September,  50-75%  of  the  sample  were  brooding,  but  low 
levels  of  about  10%  brooding  often  occurred  during  the  "off"  season.  Although 
the  brooding  frequencies  of  outfall  and  control  samples  of  C.  fissus  were  often 
quite  different  at  any  given  month,  there  was  no  consistent  difference  in  the  overall 
brooding  cycle  timing  or  amplitude  from  the  two  areas.  Broods  in  all  develop- 
mental stages  were  found  in  both  populations  at  all  times  of  the  year. 

The  control  populations  of  Balanus  gland nhi  brooded  embryos  in  winter  and 
spring  from  about  December  or  January  to  May,  with  about  60—80%  of  the  control 
population  brooding  during  peak  months.  Occasional  low  levels  (about  5%)  of 
brooding  occurred  in  fall  months,  but  the  onset  of  the  brooding  cycle  was  abrupt. 
The  samples  from  the  warm-water  outfall  consistently  had  a  lower  percentage  of 


RKL'KOnrCTIOX    IX    RARXACLES 


265 


brooding,  and  the  onset  of  brooding  \vus  delayed  one  or  more  months  in  both  1972— 
7$  and  1974-75,  but  not  in  l(>73-74.  The  samples  at  the  onset  of  the  brooding 
period  each  year  showed  a  high  frequency  (about  95%)  of  broods  in  early  develop- 
mental stages,  suggesting  that  the  deposition  of  the  first  brood  was  quite  syn- 
chronous in  both  populations.  Subsequent  samples  in  the  brooding  season  did  not 
reflect  any  synchrony,  and  all  developmental  stages  of  brooded  embryos  were  found. 


Chthamalus  fissus  i\ 


i  \ 


*-^ 


NDJFMAMJ    J   ASONDJ   FMAMJ   J   ASONDJ 


80 
cr> 
C70 


00 


o> 

o 

i_ 

CD 


40 
30 
20 
10 
0 


1974 

Ba/anus  glandula 


NDJ   FMAMJJASONDJFMAMJJASONDJ 


'72 


1973 


1974 


Tetraclita  souamosa 


NDJ 
'72| 


AMJJASONDJ    F 
1973 


AMJ    JASONDJ 
1974 


FIGURE  2.  Cycles  of  brooding  frequencies  :  outfall,  solid  symbols ;  control,  open  symbols. 
Sample  sizes  are:  Chthaiiuilns  fissus  —  50  ;  Balaints  glandula  =  60  ;  Tetraclita  sqnaniosa  =  60  ]> 
18  mm  basal  diameter. 


266 


AN  SON  H.  HINES 


The  control  population  of  Telnic/ita  si/im/nusa  brooded  during  summer  from 
about  June  through  September.  The  onset  of  brooding  was  sharp,  and  40-75% 
of  the  population  brooded  embryos  during  peak  months.  The  population  in  the 
warm-water  outfall  began  brooding  in  or  near  December  and  continued  at  errati- 
cally variable  levels  through  spring  into  early  summer,  diminishing  in  |une  or  July 
when  the  control  population  was  reaching  peak  activity.  The  brooding  cycle  of  the 
outfall  population  was  thus  about  six  months  out  of  phase  with  the  control  and 
more  variable  in  activity,  of  longer  duration  but  with  lower  peak  brooding  fre- 
quencies than  the  control  population.  Samples  from  the  control  population  at  the 
onset  of  the  brooding  season  showed  a  high  frequency  (about  80c/o )  of  broods  in 
early  developmental  stages,  indicating  a  synchrony  of  deposition  of  the  first  brood. 
This  synchrony  was  not  found  in  subsequent  samples  in  the  season,  and  it  was 
not  found  at  all  in  the  samples  of  the  outfall  population. 

Brooding  frequencies  as  a  function  of  size  were  calculated  for  each  species. 
There  was  no  significant  change  in  brooding  activity  with  size  in  Clitluiniitlns  fissns; 
any  individual  above  2  mm  basal  diameter  (about  2  months  old;  Hines,  1976)  was 
judged  to  be  mature.  No  individual  of  Bui-amis  glandnla  less  than  5  mm  basal 
diameter  was  available  for  sampling  during  months  of  brooding  activity,  since  they 
had  all  grown  to  at  least  that  size  by  December  (about  6  months  old;  Hines,  1976). 
Above  5  mm  there  was  no  significant  change  in  brooding  activity  with  size,  so  all 
individuals  were  considered  mature  by  the  time  the  populations  began  to  brood  in 
the  winter.  In  Tetniclita  synaiiiosa,  however,  barnacles  less  than  6  mm  basal  diame- 
ter did  not  brood.  They  began  to  mature  at  about  12  mm,  and  became  fully  mature 
at  about  18  mm  in  diameter  (at  about  two  years  old;  Hines,  1976).  Above  18  mm 
in  size  there  was  no  significant  change  in  brooding  activity. 

Laboratory  experiments  on  brooding 

Comparisons  of  brooding  cycles  in  the  warm-water  outfall  and  control  popu- 
lations suggested  that  for  Chthamalus  fissus  temperature  is  not  an  important  factor 
regulating  brooding,  since  the  cycles  of  the  two  populations  are  similar.  In 
Balanus  ylandnhi  the  delayed  and  lower  percentages  of  brooding  in  the  outfall 
population  suggested  that  temperature  is  important  in  regulating  both  the  timing 

TABLE  I 

Laboratory  brood-ing  experiment.  Chthamalus  fissus  and  Balanus  glaudula  collected  from  both  the 
outfall  and  control  populations  were  maintained  in  the  lab  from  October  12  to  December  30,  JQ73: 
photoperiod,  12L:12D;  food,  Artemia  nauplii  fed  ad  lib;  temperatures,  11.5°  or  20°  C.  Brooding 
frequencies  for  the  lab  barnacles  and  the  field  populations  at  the  end  of  the  experiment  are  shown. 


11.5°  C 

20°  C 

Field  12/20/73 

Control 

Outfall 

Control 

Outfall 

Control 

Outfall 

Chthamalus  fissus 

N  = 

49% 

58 

42% 
56 

46% 

38 

40% 
62 

14% 
50 

13% 
50 

Balanus  glandnla 

N  = 

7V, 
77 

61% 

76 

10% 
80 

v' 

°  /O 

80 

0% 
60 

0% 
60 

REPRODUCTION  IN   BARNACLES 


TABLE  II 


267 


Laboratory  brooding  experiment  on  Chthamalus  fissus  >n<iint<iined  in  the  laboratory  from  September 
17  to  October  28,  1974:  photoperiod,  12L.-J2D;  temperature,  12°  C.  Brooding  frequencies  as  a  func- 
tion of  increasing  food  doses  o/Artemia  salina  nanplii  are  shown. 


Food  dose  (ml) 

0 

10 

31 

58 

100 

Brooding  (%) 

15 

18 

33 

48 

74 

N 

103 

66 

98 

87 

66 

and  intensity  of  brooding,  and  that  cold  temperature  is  necessary  for  reproduction 
to  proceed  normally.  In  Tetraclita  sqitainosa  the  pronounced  shift  in  the  brooding 
cycle  of  the  outfall  population  suggested  that  temperature  is  important  in  regulating 
reproduction  in  this  species  also,  but  that  warm  temperatures  are  required  for 
reproduction.  A  series  of  laboratory  experiments  was  conducted  to  test  these 
hypotheses  and  to  investigate  more  fully  the  roles  of  temperature,  photoperiod,  and 
food  availability  in  regulating  reproduction.  Results  of  these  experiments  will  only 
be  summarized  here;  further  details  and  complete  data  are  available  from  Hines 
(1976). 

Brooding  in  Chtliauialus  fissus  is  regulated  directly  by  food  availability,  and 
feeding  with  Artemia  salina  nauplii  in  the  laboratory  elicited  high  brooding  fre- 
quencies (Table  I)  during  periods  when  brooding  activity  and  food  levels  in  the 
field  were  low  (Icanberry  and  Adams,  1974).  The  increased  brooding  response 
to  food  in  the  laboratory  was  rapid  (within  about  2  weeks),  and  the  frequency  of 
brooding  was  directly  proportional  to  the  size  of  the  food  dosage  (Table  II). 
Temperature  (11.5°  or  20°  C)  and  photoperiod  (10L:14D,  12L:12D,  or  14L: 
10D)  did  not  affect  brooding  in  C.  fissus  in  the  laboratory  during  any  season.  For 
Balanus  glandula  cold  temperature  (11.5°  C)  induced  early  brooding  in  the 
laboratory  during  late  fall  and  early  winter,  and  warm  temperatures  (20°  C) 
tended  to  inhibit  it  (Table  I).  However,  cold  temperature  in  the  laboratory  did  not 
induce  brooding  in  late  summer  to  early  fall  even  though  B.  glandula  appeared  ripe 
then ;  nor  did  cold  temperature  in  the  laboratory  extend  the  brooding  period  into 
summer.  Photoperiod  (10L.-14D,  12L:12D,  14L:10D)  did  not  affect  brooding 
activity  during  any  season  in  B.  glandula.  Although  the  6-month  shift  in  the 
brooding  cycle  of  Tetraclita  squauwsa  in  the  outfall  population  strongly  suggests 
that  warm  temperatures  stimulate  brooding,  T.  squauwsa  did  not  brood  in  the 
laboratory  under  any  of  the  conditions  tested  (combinations  of  11.5°  or  20°  C  with 
10L:  14D,  12L:  12D,  14L:  10D  and  several  food  regimes),  even  during  the  time 
the  field  populations  were  brooding.  Individuals  appeared  ripe  with  yolk  in  the 
laboratory,  but  the  stimulus  for  brooding  seemed  missing. 

Brood  and  egg  size 

Regressions  of  dry  brood  weight  vs.  dry  body  weight  with  95%  confidence  in- 
tervals for  slopes  and  intercepts  are:  C.  fissus:  y  -  :  (0.718  ±  0.039)  X  +  (0.035 
±  0.030),  n  ==  391,  r  =  +  0.855  ;  B.  glandula:  y  :  =  (1.58  ±  0.093)  X  -  (0.962  ± 
0.526).  n  ==  363;  r  =  +  0.868;  and  T.  squamosa:  y  :  =  (1.21  ±  0.066)  X  --  (8.86 
±  1.82),  n  =  248,  r  =  +  0.908.  Brood  weight  is  quite  variable  in  all  three  species; 


268  ANSON  H.  HIKES 

however,  there  \vas  no  significant  difference  in  brood  weight/body  weight  regres- 
sions between  outfall  and  control  population  or  between  broods  occurring  early  or 
late  in  the  respective  brooding  season  of  any  of  the  three  species  (slopes  and  inter- 
cepts are  not  different  at  the  0.20  level).  Because  brood  weight  is  positively  cor- 
related with  body  size  and  because  there  is  a  large  size  range  of  barnacles  both 
within  each  species  and  between  species,  the  slope  of  these  regressions  is  taken  as 
a  relative  measure  of  the  brood  size  for  each  species.  By  this  measure  Balanus 
glandula  has  the  largest  relative  brood  size  (1.58)  ;  Tctraclita  sqitaniosa  puts  out  an 
intermediate  brood  (1.21)  ;  and  Chthamalus  fissus  has  a  comparatively  small  brood 
(0.718). 

Numbers  of  eggs  in  thousands  per  brood  as  a  function  of  dry  body  weight  are 
given  in  the  following  regressions  showing  the  standard  errors  of  the  slopes  and 
intercepts:  C.  fissus:  y  =  2.54  (±0.34)  X  -- 0.28  (±0.27),  n  ==  24,  r==0.85; 
B.  glandula:  y  =  1.72  (±0.13)  X  -- 0.37  (±0.98),  n  ==  33,  r==0.92;  and  T. 
sifiianiosa:  y  ==  0.75  (±0.08)  X  --  2.05  (±1.15),  n  ==  33,  r  ==  0.87. 

Size  of  the  ovoid-shaped  eggs  of  each  species  are  given  by  their  length  and  width 
at  the  first  naupliar  stage  before  hatching :  130  X  95  /*  for  Chthauialns  fissus;  245  X 
175  }t,  for  Balanns  glandula;  and  340  X  195  ^  f°r  Tctraclita  squaiuosa.  These 
dimensions  varied  only  by  about  ±5  p.  within  each  species. 

Body  zi'cight  relative  to  opercular  ivciglit 

The  regressions  of  body  weight  relative  to  opercular  valve  weight  for  each  out- 
fall and  control  sample  were  computed.  The  intercepts  of  all  these  regressions 
were  all  near  zero,  and  changes  in  the  slopes  were  interpreted  as  measures  of 
fluctuations  in  body  weights  of  the  barnacles.  The  body  weights  of  the  three  species 
fluctuated  erratically  throughout  the  year,  but  there  was  no  discernible  seasonal 
cycle  of  body  weight  for  any  of  the  three  species  (Fig.  3).  The  control  samples 
had  consistently  higher  body  weight  to  opercular  valve  weight  ratios  than  outfall 
samples  for  all  three  species,  except  for  the  second  year  of  data  for  Balanus  glandula 
where  there  was  no  difference  between  the  two  populations.  The  body  weights  are 
not  large  compared  to  brood  weight  in  all  species,  and  it  is  hard  to  see  how  any 
sizeable  quantity  of  nutrients  could  be  stored  there.  Variability  of  body  weights  in 
monthly  samples  may  reflect  differences  other  than  stored  nutrients,  e.g.,  gut  con- 
tents. Furthermore,  differences  between  outfall  and  control  samples  could  have 
been  due  to  differences  in  opercular  valve  weights.  Because  the  opercular  valves 
are  much  heavier  than  the  bodies,  small  increases  in  calcium  deposition  at  the  higher 
temperatures  in  the  outfall  could  have  accounted  for  the  smaller  body  weight  to 
opercular  valve  weight  ratios.  The  body  weight  cycles  did  not  correlate  with  cycles 
of  male  reproductive  systems  of  any  of  the  species.  It  was  therefore  decided  that 
body  weights  do  not  provide  good  reflections  of  nutrient  storage  patterns  for  these 
three  species. 

Male  reproductive  system 

Cycles  of  the  male  reproductive  systems  of  the  three  species  are  shown  in  Figure 
4.  Nearly  all  specimens  of  Cntliaiiialiis  fissus  have  ripe  male  reproductive  tracts 


REPRODUCTION  IN  BARNACLES 


269 


0.3r 


0.2 


O.I 


0) 
Q. 

.2    o 


Chthamalus 
fissus 


GO    NDJFMAMJ JASON DJFMAMJJASON 

1973  1974 


0.3 


CD 


-g    0.2 

O 

0> 
CL 

O    O.I 


"   } 


Bo  Ion  us  glandula 


NDJFMAMJ  JASONDJFMAMJ  JASON 
!          1973  1974 


0.2 


0. 


Tetrad ita  squamosa 


NDJFMAMJ  JASONDJFMAMJJASON 
1973  1974 


FIGURE  3.  Body  weight  fluctuations.  Slopes  and  95%  confidence  limits  for  regressions  of 
body  weight  versus  opercular  valve  weight  are  plotted.  Outfall  population  is  represented  by 
solid  symbols ;  control  population,  open  symbols. 


year-round,  with  no  significant  difference  between  outfall  and  control  populations. 
Balanus  (jlanduhi  had  a  definite  cycle  of  the  male  system,  developing  in  the  fall  from 
September  to  November,  remaining  ripe  during  the  brooding  season  from  December 
to  May,  and  rapidly  becoming  quiescent  in  summer  from  June  to  August.  Both 
populations  were  very  synchronous  and  similar  each  year,  although  the  control 


270 


ANSON  H.  HINES 


100 


50 


Chthamalus  fissus 


N   D,J 


100- 


i    i     i     i     t     i     i     i i i i i i i i i I I i i i — i — I — i 

FMAMJJASONDJFMAMJJASOND 

1974 


Ba/onus 
glandula 


cr 


NDJFMAMJJASONDJFMAMJJASOND 
1973  1974 


Tetraciita 

squamosa 


NDJFMAMJJASONDJFMAMJJASOND 
|  1973  1974 

FIGURE  4.  Cycles  of  male  reproductive  systems :  outfall,  solid  symbols ;  control,  open  sym- 
bols. Sample  sizes  are:  Chthamalus  fissus  =  50',  Balanus  glandula  and  Tetraciita  squamosa  = 
60. 


population  became  ripe  slightly  in  advance  of  the  outfall  population.  In  Tetraciita 
squamosa  the  control  population  showed  a  distinct  cycle  of  the  male  system  with 
peak  activity  occurring  from  about  April  to  October  and  low  frequencies  of  ripe 
individuals  from  November  to  February,  when  development  of  the  male  tracts 
began  again.  In  contrast,  the  outfall  population  tended  to  be  at  peak  frequencies 


REPRODUCTION  IN  BARNACLES 


271 


of  ripeness  during  winter,  with  erratic  intermediate  levels  of  activity  during  the 
rest  of  the  year.  The  \\ann-\vater  outfall  obviously  had  a  major  disruptive  effect 
on  the  cycle  of  development  of  the  male  system  in  this  species. 

Ovarian  cycles  and  ovary  u'cit/lit-hody  ivcnjht  regressions 

Cycles  of  ovarian  ripeness  are  shmvn  in  Figure   5.     Ovarian  development  in 
CJitlnuiialiis  fissus  showed  the  same  cycle  as  that  of  brooding  frequency,  except  that 


80- 

70- 

60 

50 

40 

30 
20 

10 
0 


Chthamalus  fissus 


NDJ  FMAMJ  JASONDJ  FMAMJJAS(  ND 
973  1974 


\   //  Bo /anus  glondula 
\ 


CL 


°NDJFMAMJ  JASONDJ  FMAMJJASO 

1973  !  1974 

90- 
80- 


70 
60 
50 


Tetrad ita  squamosa 


NDJ  FMAMJJASONDJ  FMAMJJASOND 
1973  1974 


FIGURE  5.    Ovarian  cycles  :  outfall,  solid  symbols ;  control,  open  symbols.     Sample  sizes  are 
ChtJumialits  fissus  —  50;  Balanus  yhmdnla  and  Tctraclita  sqiuimosa  —  60. 


272 


ANSON  H.  HINES 


the  summer  peaks  in  ovary  development  did  not  reach  as  high  frequencies.  As  in 
brooding  frequencies,  there  were-  no  (.-(insistent  differences  between  outfall  and  con- 
trol cycles  of  ovarian  ripeness.  The  ovaries  of  Balaniis  ylandula  ripened  rapidly  in 
the  summer  from  June  to  August.  Nearly  the  whole  population  was  ripe  from 
August  through  December  or  January  when  brooding  began.  The  frequencies  of 
ripe  individuals  dropped  during  the  brooding  period  to  a  low  in  April  and  May. 
The  cycles  for  outfall  and  control  populations  were  nearly  identical,  except  that  the 
control  population  appeared  to  have  spent  its  ovaries  somewhat  in  advance  of  the 
outfall  population,  which  might  be  expected  since  it  began  brooding  sooner.  The 
ovarian  cycle  of  the  control  population  of  Tetraclita  sqnainosa  showed  that  these 
barnacles  became  ripe  in  the  late  spring  from  April  to  May,  slightly  in  advance  of 
the  brooding  period,  and  percentages  of  ripe  individuals  dropped  precipitously  in 
July  and  August  when  the  summer  brooding  peak  \vas  reached.  Peak  frequencies 
of  ripe  individuals  roughly  corresponded  to  the  peak  brooding  frequencies  attained. 
However,  the  warm-water  outfall  population  had  only  low  frequencies  of  ripe 
ovaries  during  a  long  period  from  October  to  May  or  June,  encompassing  the  same 
time  as  the  brooding  period.  This  population  had  virtually  no  ripe  individuals  in 
the  summer  months. 

To  estimate  the  number  of  broods  for  which  the  ovary  stored  nutrients,  ovary 
weights  were  analyzed  during  months  when  each  species  was  maximally  ripe  (June, 
1973,  for  Chthamalus  fissus;  December,  1973,  for  Balanns  ylandida;  and  June,  1974, 
for  the  Tetraclita  squaiuosa  control  population),  and  also  during  months  when  they 
were  least  ripe  (December,  1973,  for  C.  fissus;  May,  1973,  for  B.  f/Iandnla;  and 


en 
E 


Chthamalus  fissus 


unnpe_oyary_ 


8       10 


1.4 


Bo/anus  glandula 


180 


Tetraclita 

squamosa 


16  5  10         15         20 

Body  Weight   (mg) 


FIGURE  6.     Regressions  of  ripe  ovary  weight,  unripe  ovary  weight,  and  brood  weight  versus 
body  weight.    See  text  for  further  explanation. 


REPRODUCTION  IN  BARNACLES  273 

February,  1974,  for  T.  sqiiainosa  controls).  The  latter  series  of  ovary  weights 
were  used  as  a  baseline  for  "unripe  ovary  weight",  because  some  of  the  "ovary"  was 
actually  retractor  muscles  and  tissue  lining  the  mantle  chamber.  Ripe  ovaries  are 
extremely  diffuse  and  ramify  through  the  compartment  walls,  so  that  some  ovarian 
tissue  always  remained  uncollected  when  scraped  out  for  weighing.  The  regressions 
for  "ripe"  and  "unripe"  ovary  weight  versus  body  weight,  as  well  as  the  brood 
weight  versus  body  weight  regressions,  are  shown  in  Figure  6.  In  C.  fissns  the 
unripe  ovary  was  essentially  nonexistent,  and  no  tissue  could  be  collected  for  weigh- 
ing; as  the  ovary  ripened,  it  increased  up  to  a  weight  about  equal  to  the  brood  size, 
indicating  that  not  more  than  one  brood  was  prepared  at  a  time.  In  B.  glandula 
the  unripe  ovary  and  associated  tissues  were  significant  in  size  and  the  ovary 
showed  an  increase  of  about  2.5  times  the  brood  weight  when  ripe.  This  indicated 
that  yolky  material  for  a  minimum  of  three  broods  and  possibly  more  was  stored 
in  the  ovary,  considering  that  much  yolky  ovarian  tissue  was  left  behind  in  the  shell 
during  collection.  In  T.  squaiiiosa  although  both  ripe  and  unripe  ovaries  and  the 
associated  tissues  weighed  more  than  the  brood,  ripe  ovaries  were  only  slightly 
larger  than  unripe  ovaries.  Therefore,  although  these  tissues  were  potentially  size- 
able storage  areas  in  T.  squainsa,  not  much  nutrient  storage  in  the  form  of  yolky 
material  detectable  by  weight-change  actually  occurred — at  most  one  brood  wras 
prepared  in  advance. 

Time  per  brood  and  number  of  broods  per  season 

Preliminary  estimates  of  the  time  each  brood  was  retained  in  the  mantle  chamber 
were  made  for  Chthainalus  fissus  and  Balanns  glandnhi  by  keeping  barnacles  moist, 
but  not  submerged,  in  a  12°  C  cold  room.  Periodic  samples  of  barnacles  were  in- 
spected for  the  developmental  stages  of  brooded  embryos.  Since  copulation  did  not 
occur  and  no  new  broods  were  laid  down  unless  the  barnacles  were  submerged, 
only  already  existing  broods  continued  to  develop  without  hatching.  In  C.  fissns, 
100%  of  the  broods  were  judged  ready  to  hatch  in  about  two  weeks.  In  B. 
glandula,  all  broods  were  ready  to  hatch  after  one  month,  but  the  adults  were  in 
poor  condition  after  this  much  time  without  submergence.  Tetraelita  sq mimosa 
could  not  be  maintained  unsubmerged  for  any  extended  period. 

Patel  and  Crisp  (19601))  showed  that  brooding  time  determined  in  vitro  cor- 
responded well  with  in  vivo  times  for  several  species  of  barnacles.  To  get  a  better 
estimate  of  the  time  per  brood,  freshly  deposited  eggs  were  removed  from  the  mantle 
chambers  of  each  species  and  maintained  in  vitro.  C.  fissus  nauplii  did  not  hatch 
in  vitro  but  development  appeared  to  continue  normally  to  a  darkly  pigmented, 
eyed  nauplius  stage  which  was  judged  ready  to  hatch  by  comparison  with  the  most 
advanced  embryos  brooded  in  the  field.  The  reason  hatching  did  not  occur  is  not 
known,  and  whether  the  proper  stimulus  required  parental  presence,  or  some  en- 
vironmental factor,  or  both,  was  not  investigated.  At  12°  C,  C.  fissus  nauplii  were 
judged  to  be  ready  to  hatch  in  14  days  (s.d.  -  ±  3  days;  n  =-  21  broods).  At 
19°  C  development  time  for  C.  fissus  was  12  days  (s.d.  =  ±2  days;  n~15 
broods).  Embryos  of  B.  glandula  readily  developed  and  hatched  /';/  vitro.  At  12° 
C  time  to  hatching  was  27  days  (s.d.  =  ±  6  days;  n  —  24  broods),  and  at  19°.  C 
B.  glandula  hatched  in  22  days  (s.d.  =  ±  4  days;  n--  IS  broods).  T.  sqiianiosa 


274 


ANSON  H.  HINES 


TAHI.R  1 1 1 

Reproductive  effort  estimated  as  the  yearly  weight  allocation  to  egg  production.  Length  of  brooding  sea- 
son divided  by  the  incubation  time  per  brood  is  the  maximum  number  of  broods  per  year.  The  num- 
ber of  broods  per  year  times  the  slope  of  the  brood  weight  versus  body  weight  regression  is  brood  weight/ 
body  weight  per  year. 


Chlhamalns  fissus 

Balanus  glandula 

Telraclila  squamosa 

Length  of  season 

8  months 

()  months 

4.5  months 

Time  per  brood 

0.5  month 

1  month 

1.5  months 

Maximum  number  broods  per  year 

16 

6 

3 

Brood  wt/body  wt 

0.718 

1.58 

1.21 

Brood  wt/body  wt  per  year 

11.49      , 

9.48 

3.63 

nauplii  hatched  in  40  days  (s.d.  =  ±  8  days;  n  -  20  broods)  at  12°  C  and  30  days 
(s.d.  -  ±7  days;  n  =-  13  broods)  at  19°  C.  In  summary,  "normal  brooding  time" 
(i.e.,  at  12°  C)  was  estimated  at  about  0.5  months  for  C.  fissus  (Qw  -  1.25)  ;  1.0 
month  for  B.  glandula  (Qio  =  1.34);  and  1.5  months  for  T.  squamosa  (do  = 
1.51). 

To  calculate  a  maximum  number  of  broods  per  year  produced  by  the  control 
populations  of  each  species,  the  length  of  the  brooding  season  was  divided  by  the 
estimated  time  required  for  each  brood,  yielding  about  16  broods  per  season  for 
Chthamalus  fissus,  six  for  Balanus  glandula,  and  three  for  Tetraclita  squajiiosa  (see 
Table  III).  This  calculation  makes  the  assumption  that  there  is  no  delay  between 
broods.  Probably  at  least  a  short  period  between  broods  in  fact  occurs,  because  less 
than  100%  of  the  barnacles  were  brooding  at  any  given  time.  However,  there  is 
strong  circumstantial  evidence  that  each  species  puts  out  several  broods  in  rapid 
succession,  as  shown  by  Crisp  and  Davies  (1955)  and  Patel  and  Crisp  (1960a,  b) 
for  other  species.  The  length  of  the  brooding  season  for  each  species  in  the  present 
study  was  clearly  much  longer  than  the  developmental  time  for  each  brood,  and 
most  barnacles  had  ripe  ovaries  remaining  while  still  brooding.  Embryos  of  all 
stages  of  development  were  found  throughout  the  brooding  season  of  each  species, 
and  occasionally  a  new  brood  was  present  while  advanced  embryos  from  a  previous 
brood  remained.  Thus,  the  assumption  that  there  was  minimal  delay  between 
broods  appears  reasonable. 

Any  delay  between  broods  would  tend  to  decrease  the  number  of  broods  pro- 
duced in  a  season.  For  example,  a  lag  of  only  4—5  days  between  broods  in  Balanus 
glandula  would  reduce  from  six  to  five  the  estimate  of  the  number  of  broods  pro- 
duced over  the  6-month  season,  and  this  short  lag  could  result  in  15-20%  of  the 
population  not  brooding  at  any  given  time.  On  the  other  hand,  any  factor,  such 
as  temperature  higher  than  12°  C,  which  shortens  the  brooding  time  would  tend  to 
make  the  production  of  more  broods  possible.  It  is  thus  very  difficult  to  estimate 
the  number  of  broods  produced  by  the  outfall  populations.  B.  glandula,  for  ex- 
ample, tends  to  have  a  delayed,  shorter  brooding  season  in  the  outfall  with  a  lower 
percentage  of  the  population  brooding  at  any  given  time,  but  the  higher  tempera- 
tures probably  shorten  the  time  each  brood  is  retained.  B.  glandula  in  the  control 
population  thus  produces  a  minimum  (based  on  storage)  of  three  and  a  maximum 
(based  on  brooding  time)  of  six  broods  per  season.  It  is  difficult  to  estimate  a 


REPRODUCTION  IN  BARNACLES 


275 


minimum  number  of  broods  per  season  for  Chthamalus  fissus  and  Tetraclita  sqita- 
inosa,  because  they  do  not  store  yolk  for  more  than  one  brood  at  a  time. 

Patterns  oj  nutrient  storage  and  reproductive  effort 

A  schematic  model  of  the  patterns  of  nutrient  storage  during  the  year  is  pre- 
sented in  Figure  7,  showing  proportional  changes  in  ovary  size  relative  to  brood 
size  for  each  species.  Broods  in  Chthamalus  fissus  are  deposited  as  soon  as  enough 
nutrients  are  accumulated.  As  plankton  production  increases  during  spring  and 


CD 


a 

O 

c: 


o 

Z3 

cr 


o 
o 

^ 

CD 


O 

o" 


0 


Chthamalus  fissus 


g1         DJFMAMJJASOND 


0 


Balanus  glandula 


o 


DJ      FMAMJ      J      ASOND 


Tetraclita  squamosa 


DJFMAMJJASOND 


FIGURE  7.     Schematic  model  of  patterns  of  nutrient  storage  in  ovary.     Amount  of  yolky 
material  equivalent  to  the  weight  of  a  brood  is  shown  over  a  one  year  cycle  for  each  species. 


276  ANSON  H.  HINES 

summer,  the  rate  of  yolk  accumulation  increases  and  broods  are  produced  more 
frequently,  until  brood  production  is  limited  by  the  time  required  for  brood  incuba- 
tion. In  contrast,  Balanus  (jlandula  rapidly  stores  nutrients  for  at  least  three 
broods  during  the  summer  and  remains  ripe  until  cold  temperatures  induce  brood- 
ing. As  broods  are  released  during  winter  and  spring,  more  nutrients  are  prob- 
ably added  as  the  barnacles  feed  until  they  are  spawned  out  in  May  after  up  to  six 
broods.  Tctraclita  stjuaniosa  produces  only  three  broods  in  the  summer,  and  only 
yolk  for  one  brood  at  a  time  is  accumulated. 

Comparisons  of  the  reproductive  efforts  of  the  three  species  estimated  as 
brood  weight  relative  to  body  weight  per  year  are  shown  in  Table  III.  For  each 
species  the  estimated  maximum  number  of  broods  produced  per  season  is  multi- 
plied by  the  slope  of  the  regression  of  brood  weight  on  body  weight  to  calculate  the 
total  size-specific  brood  weight  expended  during  the  year.  Chthainalus  fissns  has 
the  highest  proportional  reproductive  effort — 11.49  times jthe  body  weight  produced 
as  broods;  Bahnuis  (jlandula  expends  an  intermediate  but  large  amount — 9.48  times 
the  body  weight ;  and  Tctraclita  sijitainosa  has  the  lowest  weight  allocation  to  brood 
production — 3.63  times  the  body  weight. 

DISCUSSION 

Species  of  barnacles  may  be  grouped  into  roughly  five  categories  based  on  a 
spectrum  of  reproductive  patterns :  first,  a  boreo-arctic  pattern  in  which  a  single 
large  brood  is  incubated  over  winter ;  secondly,  a  pattern  shown  by  a  few  cold- 
temperature  species  which  produce  a  small  number  of  broods  in  winter  and  spring; 
thirdly,  a  variable  pattern  found  in  several  warm-temperature  and  subtropical 
species  which  produce  many  small  broods  during  summer ;  fourthly,  a  possibly  dif- 
ferent summer  pattern  of  producing  only  a  few  broods,  as  demonstrated  here  by 
Tetraclita  squanwsa;  and  fifthly,  a  pattern  of  brooding  throughout  the  year. 
Some  species  may  show  different  patterns  of  reproduction  in  different  parts  of  their 
geographic  range. 

The  boreo-arctic  species  Balanus  hameri,  B.  balanus,  and  especially  B.  balanoidcs 
are  perhaps  the  best  understood  and  have  the  most  precisely  timed,  least  complex 
pattern  of  brood  production.  These  barnacles  store  nutrients  during  a  refractory 
summer  period  when  brooding  cannot  be  induced  (Barnes,  1963).  In  B.  balanoidcs 
copulation  occurs  in  late  fall,  cued  by  a  low  temperature  threshold  modified  by 
photoperiod  (Barnes,  1963),  and  in  B.  balanus  and  B.  Jiaiucri  copulation  occurs  in 
mid-winter  (Crisp,  1954;  Barnes  and  Barnes,  1954).  A  single  large  brood  is  in- 
cubated over  winter,  and  naupliar  release  is  synchronized  with  the  spring  diatom 
bloom  by  chemical  agents  produced  by  the  parents  and/or  the  diatoms  (Crisp, 
1956;  Barnes,  1957).  Clearly,  the  adaptive  significance  of  this  precise  timing  is 
based  on  the  predictability  of  the  marked  seasonal  changes  in  temperature,  photo- 
period,  and  productivity  of  northern  latitudes. 

Verruca  strocutia  and  Balanus  c/Iandiila  in  cold  temperate  waters  exemplify  a 
different  pattern  of  brood  production  during  winter  and  spring.  V.  strocinia  is 
similar  to  the  borea-arctic  species  in  that  it  produces  a  major  brood  in  the  winter 
which  is  synchronous  throughout  the  whole  population,  and  which  is  released  about 


REPRODUCTION  IN  BARNACLES  277 

the  time  of  the  spring  diatom  increase  (  P>arnes  and  Stone,  1973 J.  However,  unlike 
the  boreo-arctic  species,  minor  broods  are  subsequently  produced  asynchronously 
during  spring  and  summer.  Barnes  and  Barnes  (1956)  found  brooding  cycles  for 
B.  glandiila  at  British  Columbia  and  southern  California  similar  to  that  reported 
here  for  central  California.  The  northern  population  showed  a  shorter  period  of 
high  brooding  frequency  from  January  to  March,  with  low  frequencies  occurring 
erratically  during  spring  and  summer,  while  their  southern  population  showed  a 
broader  season  of  high  brooding  frequencies  from  November  to  May.  The  length 
of  the  season  reported  here  for  Morro  Bay  is  intermediate  from  December  or  Janu- 
ary to  May.  Since  the  present  laboratory  experiments  and  the  delay  of  brooding  in 
the  warm-water  outfall  clearly  show  that  the  initiation  of  brooding  in  B.  ylandula 
is  regulated  by  cold  temperature,  the  longer  brooding  season  of  the  southern  popu- 
lations is  hard  to  explain.  The  brooding  season  of  the  Morro  Bay  population  was 
not  extended  with  laboratory  manipulations.  In  any  case,  based  on  cycles  of 
brooding  frequency,  Barnes  and  Barnes  (1956)  suggest  B.  glandiila,  like  V .  stroc- 
111  in.  produces  a  single  major  winter  brood  followed  by  a  second  minor  brood  in 
spring.  At  Morro  Bay  at  least  three  and  as  many  as  six  broods  are  produced,  with 
the  first  being  nearly  synchronous  in  winter  and  the  others  following  at  about 
monthly  intervals  through  spring.  This  pattern  apparently  times  the  settlement  of 
many  larvae  in  late  spring  and  early  summer,  when  warming  temperatures  and  high 
food  availability  are  optimal  for  growth  (Hines,  1976),  and  this  pattern  distributes 
the  chance  of  reproductive  success  over  several  broods. 

Several  species  of  cirripedes  in  warm-temperate  and  subtropical  regions  char- 
acteristically produce  numerous  small  broods  in  rapid  succession  during  summer. 
The  length  of  the  brooding  season  is  usually  broadly  defined  by  temperature,  and 
during  the  season  the  production  of  broods  is  only  limited  by  food  availability  for 
restoring  the  ovary  and  by  the  temperature-dependent  development  rate  of  the 
brood  in  the  mantle  chamber  (Crisp,  1950;  Patel  and  Crisp,  1960a,  b).  The  incu- 
bation time  per  brood  as  a  function  of  temperature  has  been  measured  for  a  number 
of  species  (Patel  and  Crisp,  1960b),  allowing  a  calculation  of  about  13-22  broods 
produced  in  Chthamalus  stcllatus  and  10-25  broods  in  Balanus  amphitrite  dcntic- 
ulata  during  spring  and  summer.  C.  fissus  in  the  present  study  produces  about  16 
broods.  Although  its  brooding  season  is  limited  by  food  availability  rather  than 
temperature,  its  reproductive  pattern  fits  in  this  category.  This  seasonal  pattern  is 
adaptive  for  quick,  opportunistic  response  to  short-term  changes  in  the  environ- 
ment, while  minimizing  the  energtic  cost  of  any  single  brood  if  conditions  turn  bad. 
It  also  disperses  sibling  larvae  and  improves  chances  of  colonization  (Strathmann. 
1974). 

In  the  present  study  Tctraclita  siinainosa  produces  only  about  three  broods  in 
summer,  each  incubated  for  a  long  period.  This  is  contrary  to  the  pattern  in  which 
those  species  that  reproduce  in  summer  tend  to  produce  numerous,  frequent  broods, 
while  those  that  put  only  a  few  broods  seem  to  be  northern  species  which  breed  in 
winter  and/or  spring.  The  early  reproduction  of  T.  squainosa  in  the  warm-water 
outfall  strongly  suggests  that  brooding  is  cued  by  warm  temperature,  although 
further  study  is  needed  since  brooding  of  this  species  could  not  be  induced  in  the 
laboratory.  T.  sqnaiiiosa  grows  to  a  size  large  enough  to  attain  immunity  from 


278  ANSON  H.  HINES 

many  predators  and  compete  successfully  for  space  at  low  tidal  levels  (Hint's, 
19/6).  Jt  does  not  become  reproductive!}'  mature  until  it  has  grown  to  about  18 
mm  in  diameter,  and  brooding  occurs  during  the  productive  summer  period. 
This  would  minimize  the  amount  of  energy  diverted  from  growth  to  reproduction 
during  critical  periods  of  small  size  and  low  food  availability. 

Elminius  modes/us  and  Balanns  pacific  us  brood  continuously  year-round,  and 
both  species  grow  and  become  reproductively  mature  very  quickly  (Crisp  and 
Davies,  1955;  Hurley,  1973).  E.  modestiis  produces  a  brood  about  every  two 
weeks  during  summer;  but  the  rate  of  brood  production  is  markedly  reduced  during 
the  cold  winter  months  when  food  availability  is  low,  resulting  in  about  12-20 
broods  per  year.  B.  pad  fie  us  produces  about  23-33  broods  per  year  with  no 
obvious  seasonal  cycle.  Both  species  are  characterized  as  colonizers,  with  E. 
inodestus  undergoing  a  rapid  range  expansion  since  its  introduction  in  Europe  and 
B.  padficus  settling  subtidally  on  newly  bared  substrate  (Crisp  and  Davies,  1955; 
Hurley,  1973). 

Some  species  of  cirripedes  with  wide  latitudinal  distributions  exhibit  variable 
reproductive  cycles.  For  example,  Poll-id  pcs  polymer  us  shows  a  variable  reproduc- 
tive season  from  Washington  to  southern  California.  The  northern  populations 
brood  in  summer  (Lewis,  1975),  and  the  length  of  the  season  increases  to  the  south 
in  northern  and  central  California  (Hilgard,  1960;  C.  Hand,  J.  Standing  and  J. 
Rutherford,  personal  communication).  Farther  south  at  Morro  Bay  brooding  oc- 
curs at  erratically  high  frequencies  year-round  (Hines,  unpublished),  and  at  Santa 
Barbara  in  southern  California  peak  brooding  activity  occurs  in  the  winter  with  at 
least  a  30%  brooding  frequency  in  the  summer  (Straughan,  1971).  Balanus 
amphitrite  denticulata  has  a  brooding  season  limited  by  temperatue  from  June  to 
August  in  Great  Britain  (Patel  and  Crisp,  1960b ),  while  the  brooding  of  B.  amphi- 
trite communis  in  India  appears  restricted  from  September  to  June  by  salinity  and 
food  availability  (Pillay  and  Nair,  1972).  Balanus  crenatus  in  northern  British 
waters  has  a  boreo-arctic  pattern  of  producing  a  single  large  winter  brood  (Barnes 
and  Barnes,  1968),  but  it  may  produce  a  second,  spring  brood  elsewhere  in  Great 
Britain  (Patel  and  Crisp,  1960b),  while  subtidal  populations  in  central  California 
produce  numerous  small  broods  year-round  with  no  obvious  seasonal  cycle  (per- 
sonal observations).  These  species  with  variable  reproductive  cycles  deserve  much 
more  study,  because  they  may  provide  valuable  insights  into  the  way  proximal 
environmental  cues  regulate  reproduction  under  different  conditions. 

Comparisons  of  reproductive  effort  in  barnacles  require  estimates  of  both  the  size 
as  well  as  number  of  broods  produced  per  year.  Because  published  data  on  brood 
weights  for  barnacles  have  not  been  available  before  the  present  paper,  Barnes  and 
Barnes  (1968)  compared  relative  brood  volumes  of  a  variety  of  barnacle  species  as 
a  next  best  indicator  of  brood  size.  For  each  species  they  calculate  a  product  (NY) 
as  a  measure  of  the  brood  volume  relative  to  the  size  of  the  barnacle.  ("N"  is  the 
increase  in  numbers  of  eggs  per  brood  per  50  micrograms  dry  body  weight  and 
"V"  is  the  volume  in  10~G  ml  of  an  ellipsoid  calculated  from  the  length  and  width 
of  the  "egg"  at  the  first  naupliar  stage.)  Brood  sizes  of  barnacles  from  their  calcu- 
lations fall  into  three  categories:  first,  boreo-arctic  species  (including  B.  balanoides, 
B.  balaniis,  and  B.  crenatus}  with  very  large  broods,  NV  —  1500-3500;  secondly, 


REPRODUCTION  IN  BARNACLES  279 

temperate  and  subtropical  species  (including  five  species  of  Balanns,  two  species  of 
Chthamahts,  and  one  each  of  Octoments,  Tctraclita,  and  FJminins)  with  inter- 
mediate but  much  smaller  broods,  NY  :  100-500;  and  thirdly,  a  few  species  (in- 
cluding two  species  of  Pollicipes,  one  Chthamalns  and  one  Verruca}  with  very 
small  broods,  NY  =  30-60.  From  their  data,  B.  glandula  (XV  =  292)  is  included 
in  the  second  group  and  C.  fissus  (XV  ;  47)  is  in  the  third  ;  no  data  are  given  for 
T.  sqnamosa.  From  the  regression  of  egg  numbers  per  brood  versus  dry  body 
weight  and  from  the  dimensions  of  the  eggs  for  the  Morro  Bay  populations,  com- 
parable NY  values  can  be  calculated  for  C.  fissus  (61)  and  B.  glandula  (265),  and 
an  additional  value  for  T.  sqitamosa  (199). 

Barnes  and  Barnes  (1968)  proposes  the  product  (NYTB)  of  the  number  of 
broods  produced  per  year  (B)  times  the  NV  value  of  a  species.  They  suggest  that 
since  boreo-arctic  species  produce  only  a  single  brood  and  warm-water  species 
many  broods,  NYB  values  are  roughly  equal  for  all  species  and  the  "metabolic 
efficiency  of  egg  production"  in  barnacles  is  constant.  However,  without  measure- 
ments of  the  amount  of  food  available  to  different  species  and  calculations  of  their 
assimilation  rates  and  energy  budgets,  the  term  "metabolic  efficiency"  is  misleading, 
and  "reproductive  output"  or  "reproductive  effort"  are  better  terms  to  describe  the 
data.  The  reproductive  efforts  of  the  Morro  Bay  populations  computed  similarly  as 
a  function  of  volume  are:  NYB  --  979  for  16  broods  in  Chthanialus  fissus;  NYB  = 
1589  for  six  broods  and  795  for  three  broods  in  Balanns  ylanditla ;  and  NYB  = 
597  for  three  broods  in  Tetraclita  sqitamosa.  By  this  measure  B.  glandula  has  the 
largest  reproductive  output  of  the  three  species.  This  ranking  does  not  cor- 
respond with  the  estimates  of  yearly  weight  allocation  to  eggs  presented  here,  al- 
though at  only  three  broods  per  year  B.  glandula  would  be  intermediate  between 
C.  fissus  and  T.  squamosa.  Moreover,  neither  these  NYB  values  nor  the  estimates 
of  reproductive  effort  based  on  weight  support  the  hypothesis  that  the  relative  re- 
productive output  of  barnacles  is  constant.  Although  the  NYB  value  for  B. 
glandula  at  six  broods  per  year  is  in  the  low  range  of  values  for  boreo-arctic 
species,  C.  fissus  would  have  to  produce  eight  and  T.  squamosa  four  to  five  addi- 
tional broods  per  year  to  approach  the  NYB  values  of  B.  balanus  or  B.  crcnatus. 
In  fact,  the  constancy  of  reproductive  effort  proposed  for  barnacles  would  not  be 
expected  (see  Stearns,  1976).  In  species  which  produce  many  small  broods 
small  NYB  variations  would  be  equivalent  to  a  difference  of  a  brood  or  two,  and 
this  may  be  very  significant  ecologically.  The  unanswered  critical  questions  center 
on  why  boreo-arctic  species  have  such  a  very  large  reproductive  output  and  on  the 
environmental  causes  for  small,  but  important  variations  in  reproductive  effort  in 
barnacles. 


This  work  formed  part  of  a  doctoral  dissertation  submitted  to  the  Department 
of  Zoology,  University  of  California,  Berkeley.  I  thank  Drs.  Ralph  I.  Smith,  Cadet 
Hand,  and  John  S.  Pearse  for  support  and  advice  during  the  study.  John  Cornell. 
Bruce  Hargreaves,  Brian  Jennison.  Margaret  Race,  James  Rutherford,  John  Sim- 
mons, Jon  Standing,  Christopher  Tarp,  lohn  Warrick,  and  an  anonymous  reviewer 


280  ANSON  H.  HINES 

helped  in  many  ways.  The  Pacific  Gas  and  Electric  Company  gave  generously  of 
time  and  facilities.  These  studies  were  funded  by  National  Science  Foundation 
Grant  GI-34932  to  Drs.  George  Trezek  and  Virgil  Schrock  of  the  Department  of 
Engineering  University  of  California,  Berkeley;  Sea  Grant  NOAA  04-5-158-20 
to  Drs.  Ralph  T.  Smith  and  Cadet  Hand  of  the  Department  of  Zoology;  and  a  grant 
from  the  Pacific  Gas  and  Electric  Company.  My  wife,  Linda,  deserves  special 
thanks. 

SUMMARY 

1.  The  reproductive  cycles  and  brood  production  of  Clifhaiualits  fissits,  Balanus 
glandula  and  Tctraclita  squamosa  from  central  California  are  compared.     C.  fissits 
produces  about   16  small  broods  from  March  through  October.  B.  glandula  pro- 
duces three  to   six   relatively  large  broods  from   December  or   January   to   May. 
T.    squamosa    incubates    only    about    three    intermediate-sized    broods    from    June 
through  September. 

2.  Brooding  in  C.  fissits  is  regulated  by  food  availability,  and  yolk  for  no  more 
than  one  brood  is  stored  at  a  time.    Feeding  in  the  laboratory  elicited  high  brooding 
frequencies  during  periods  when  brooding  activity  and  food  levels  in  the  field  were 
low,  and  the  frequency  of  brooding  was  directly  proportional  to  the  size  of  the  food 
dosage.     Temperature  and  photoperiod   did   not  affect  brooding  frequencies.     B. 
glandnla  rapidly  stores  nutrients  in  the  ovary  for  about  three  broods  during  sum- 
mer.   Cold  temperatures  induce  early  brooding  in  the  laboratory  during  late  fall  and 
early  winter,  and  the  population  in  the  warm-water  outfall  showed  delayed  and 
lower  brooding  frequencies.     Photoperiod  did  not  affect  brooding  in  B.  glandula. 
T.  sqitaiuosa  in  the  warm-water  outfall  brooded  six  months  earlier  than  the  control 
population,  suggesting  warm  temperatures  are  required  for  reproduction.    Yolk  for 
only  one  brood  at  a  time  is  stored  in  T.  squamosa. 

3.  Comparisons  of  reproductive  efforts  estimated  as  brood  weight  relative  to 
body  weight  per  year  show  that  C.  fissits  has  proportionally  the  largest  brood  pro- 
duction ;   B.   glandula   an   intermediate   but   large   amount;    and    T.   squamosa   the 
smallest  reproductive  output. 

4.  It  is  proposed  that  species  of  barnacles  may  be  grouped  into  five  categories 
based  on  major  patterns  of  reproductive  timing  and  brood  production.     The  three 
species  in  the  present  paper  show  three  of  these  patterns.     The  reproductive  effort 
of  these  three  species  is  compared  with  other  cirripecles. 

LITERATURE  CITED 

BARNES,  H.,  1957.     Processes  of  restoration  and  synchronization  in  marine  ecology:    the  spring 

diatom  increase  and  the  spawning  of  the  common  barnacle,  Balanus  balanoides.  Anncc 

BloL,  33 :  67-85. 
BARNES,  H.,  1963.    Light,  temperature,  and  breeding  of  Balanus  halanoldcs.   J.  Mar.  Blol.  Assnc. 

U.K.,43:  717-728. 
BARNES,  H.,  AND  M.  BARNES,  1954.     The  general  biology  of  Balanus  balanus   (L.)    da  Costa. 

Olkos,  5 :  63-76. 
BARNES,  H.,  AND  M.  BARNES,  1956.    The  general  biology  of  Balanus  glandula  Darwin.  Pac.  Sci., 

10:  415-422. 
BARNES,  H.,  AND  M.  BARNES,  1968.     Egg  numbers,  metabolic  efficiency  of  egg  production  and 


REPRODUCTION  IN   BARNACLES  281 

fecundity:    local  and  regional   variations   in  a  number  of  common  cirripedes.     /.  Exp. 

Mar.  Bi'ol.  EcoL,  2:  135-153. 
BARNES,  H.,  AND  R.  STOXE,  1973.     The  .general  biology  of  I'erruca  strocinia    (O.  F.  Mullet). 

II.   Reproductive  cycle,   population   structure,   and   factors   affecting   release  of  nauplii. 

/.  E.vp.  Mar.  Biol.  EcoL,  12  :  279-297. 
CONNELL,    J.,    1970.      A    predator-prey    system    in    the    marine    intertidal    region.      I.    Balanns 

glandnla  and  several  predatory  species  of  Thais.    EcoL  Mono</r.,  40:  49-78. 
CRISP,  D.  J.,  1950.     Breeding  and  distribution  of  Chthainalns  stellatus.    Nature,  166:  311-312. 
CRISP,  D.  J.,   1954.     The  breeding  of  Balanns  porcatus   (da  Costa)    in  the   Irish   Sea.  /.  Mar. 

Biol.  Assoc.  U.K.,  33  :  473-494. 
CRISP,   D.   J.,    1956.     A   substance   promoting   hatching   and   liberation   of   young   in   cirripedes. 

Nature,  178:  263. 
CRISP,   D.   J.,   AND   P.   A.   DAVIES,    1955.     Observations   in   vivo   on   the   breeding   of   Elininins 

inodcstus  grown  on  glass  slides.     /.  Mar.  Biol.  Assoc.  U.K.,  34:  357-380. 
DAYTON,   P.   K.,   1971.     Competition,   disturbance,   and  community  organization:    the   provision 

and  subsequent  utilization  of  space  in  a  rocky  intertidal  community.     EcoL  Monogr., 

41 :  351-389. 
HILGARD,  G.  H.,  1960.     A  study  of  reproduction  in  the  intertidal  barnacle,  Mitella  polymcrus, 

in  Monterey  Bay,  California.    Biol.  Bull.,  119:   169-188. 
HIXES,   A.,   1976.     Comparative   reproductive  ecology  of  three   species   of  intertidal   barnacles. 

Ph.D.  dissertation.  University  of  California,  Berkeley,  259  pp.   (University  Microfilms/ 

Dissertation  Abstracts  No.  77-4480.) 
HURLEY,  A.  C.,   1973.     Fecundity  of  the  acorn  barnacle  Balanns  pacificus  Pilsbry:    a  fugitive 

species.     LimiwI.  Oceanogr.,  18:  386-393. 
ICANBERRY,  J.,  AND  J.  ADAMS,  1974.     Zooplaiikton  studies.     Pages  135-153  in  J.  R.  Adams  and 

J.  F.  Hurley,  Eds.,  Environmental  Investigations  at  Diablo  Canyon:   1972-1973.    Pacific 

Gas    and    Electric    Company,    Department    of    Engineering    Research,     San    Ramon, 

California. 
LEWIS,   C.  A.,   1975.     Reproductive  biology  and  development  of  the  gooseneck  barnacle,   Pol- 

licipes   polymcrus.   with   special   emphasis   on   peristaltic   constrictions   in   the   fertilized 

egg.     Ph.D.  dissertation.  University  of  Alberta,  Edmonton,   Canada,  320  pp. 
PATEL,   B.,   AXD   D.   J.   CRISP,   1960a.     The   influence   of  temperature   on   the   breeding   and  the 

moulting  activities  of  some  warm-water  species  of  operculate  barnacles.    /.  Mar.  BioL 

Assoc.  U.K.,  39:  667-680. 
PATEL,  B.,  AND  D.  J.  CRISP,  1960b.     Rates  of  development  of  the  embryos  of  several  species  of 

barnacles.    Physiol.  ZooL,  33:  104-119. 
PILLAY,  K.  K.,  AND  N.  B.  NAiR,  1972.     Reproductive  biology  of  the  sessile  barnacle,  Balanus 

arfiphitrite  communis  (Darwin),  of  the  south-west  coast  of  India.    Indian  J.  Mar.  Set., 

1 :  8-16. 

RICKETTS,  E.,  AND  J.  CALVIN,  1968.     Between  Pacific  tides.    Stanford  University  Press,  614  pp. 
STEARNS,  S.,  1976.     Life  history  tactics:  a  review  of  the  ideas.    Q.  Rev.  Biol.,  51  :  3-47. 
STRATHMANN,    R.,    1974.      The    spread    of    sibling    larvae    of    sedentary    marine    invertebrates. 

Am.  Nat.,  108:  29-44. 
STRAUGHAN,  D.,  1971.     Breeding  and  larval  settlement  of  certain  intertidal  invertebrates  in  the 

Santa  Barbara   Channel  following  pollution  by  oil.     Pages  223-244  in   D.   Straughan, 

Ed.,  Biological  and  oceanographic  survey  of  the  Santa  Barbara  Channel  oil  spill  1969- 

1970,    Vol.   I,   Biology    and   bacteriology.      Alan    Hancock    Foundation,    University    of 

Southern  California. 


Reference:  Biol.  Bull.,  154:  282-291.    (April,  1978) 


ADAPTATIONS  TO  INTERTIDAL  DEVELOPMENT :  STUDIES 
ON  NASSARIUS  OBSOLETUS  x 

JAN  A.  PECHENIK  2 
Woods  Hole  Oceanographic  Institution,  }\'oods  Hole,  Massachusetts  02543 

Egg  capsules  of  marine  invertebrates  are  often  described  as  "protective"  (Car- 
riker,  1955;  Hunt,  1966;  Mileikovsky,  1971;  Sverdrup,  Johnson  and  Fleming, 
1942),  but  tbe  question  of  wbat  the  capsules  protect  against  has  rarely  been  ad- 
dressed. The  possibility  that  the  egg  capsules  of  the  intertidal  mud  snail,  Nassarius 
obsoletus,  evolved  as  an  adaptation  to  reproduction  in  the  intertidal  zone  is  con- 
sidered in  this  study. 

Although  many  marine  gastropod  and  polychaete  species  deposit  egg  capsules 
or  egg  masses  in  the  intertidal  zone,  the  consequences  of  intertidal  development  to 
the  developing  young  have  not  been  examined.  Houbrick  (1973,  p.  883)  noted 
that  the  egg  masses  of  Ccnthinin  variable  are  "frequently  exposed  to  sun  and  air 
during  low  tide,"  and  "appear  resistant  to  desiccation."  Similarly,  the  jelly  masses 
of  the  polychaete  Marphysa  are  presumed  to  protect  embryos  from  the  sun  at  low 
tide  (Aiyar,  1931),  and  the  egg  masses  of  Littorina  littoralis  (--  L.  obtusata)  are 
said  to  protect  developing  embryos  from  desiccation  (Fretter  and  Graham,  1962, 
pp.  389-390),  but  data  in  support  of  these  suppositions  are  not  given.  Egg  masses 
of  the  salt  marsh  pulmonate  Melainpits  bidentatus  can  be  desiccated  for  long  periods 
of  time,  apparently  without  interference  to  embryonic  development  (Holle  and 
Dineen,  1957).  Whether  the  egg  mass  protects  the  embryos  or  whether  the  em- 
bryos themselves  are  especially  tolerant  of  desiccation  was  not  determined. 

Spawning  in  the  intertidal  zone  does  not  necessarily  imply  that  developmental 
stages  are  resistant  to  intertidal  stresses.  Several  authors  have  suggested  that 
encapsulated  invertebrate  embryos  are  susceptible  to  desiccation  (Spight,  1975; 
Kohn,  1961)  and  osmotic  stress  (Carriker,  1955;  Gibbs,  1968).  The  walls  of  egg 
capsules  from  Urosalpinx  cinerca  are  said  to  be  freely  permeable  to  a  variety  of 
organic  and  inorganic  solutions,  organic  salts,  and  dyes  (Carriker,  1955;  Galtsoff, 
Prytherch  and  Engle,  1937). 

Nassarius  obsoletus  deposits  fertilized  eggs  in  capsules  affixed  to  firm  sub- 
strates in  the  parental  habitat  (Dimon,  1905).  The  capsules,  figured  by  Scheltema 
(1962),  are  approximately  1.5  mm  high  and  contain  30  to  several  hundred  eggs 
(Costello  and  Henley,  1971).  The  encapsulated  embryos  develop  in  the  intertidal 
zone  for  approximately  one  week,  after  which  time  veliger  larvae  emerge  from  the 
capsules  to  continue  their  development  in  the  plankton  for  at  least  several  more  weeks 
before  metamorphosing  to  the  benthos  (Scheltema,  1962).  While  encapsulated, 
embryos  are  potentially  exposed  to  intertidal  stresses,  the  most  obvious  of  which  is 
desiccation.  Kanwisher  (1957)  reports  that  relative  humidities  of  40%  are  com- 

1  Contribution  number  3923  from  the  Woods  Hole  Oceanographic  Institution. 
-  Present  address :    Graduate  School  of  Oceanography,  University  of  Rhode  Island,  King- 
ston, Rhode  Island  02881. 

282 


ttASSARIUS  INTERTIDAL  DEVELOPMENT  283 

nionly  observed  in  the  intertidal  zone.  Successful  development  in  the  intertidal 
zone  may  require  egg  capsules  which  retard  water  loss  at  low  tide,  embryos  capable 
of  tolerating  extensive  dehydration,  or  preferential  placement  of  the  capsules  in 
high-humidity  situations.  Alternatively,  substantial  pre-hatching  mortality  may 
occur.  These  possibilities  were  considered  through  determinations  of  desiccation 
tolerances  of  encapsulated  N.  obsoletus  embryos,  rates  of  water  loss  from  Ar. 
obsoletus  egg  capsules  relative  to  rates  of  water  loss  from  capsules  of  the  subtidal 
species  A',  trii'ittatus,  and  studies  of  the  adult  egg-laying  behavior  of  Ar.  obsoletus. 

MATERIALS  AND  METHODS 

Egg  capsules  of  Nassarhts  obsoletus  were  removed  from  Fiicus  and  eel  grass 
collected  at  the  Barnstable  mudflats  on  Cape  Cod,  Massachusetts.  Egg  capsules  of 
A^.  triz'ittafiis  were  deposited  in  the  laboratory  by  adults  dredged  from  Buzzard's 
Bay,  Massachusetts.  Undamaged  capsules  of  both  species  were  sorted  into  two  age 
groups  before  each  experiment,  based  on  the  extent  of  anatomical  differentiation  of 
the  enclosed  embryos  visible  at  a  magnification  of  25  X.  Capsules  containing 
"early"  embryos  (no  velum  pigmentation  or  shell  visible)  were  distinguished  from 
those  containing  "advanced"  embryos  (distinct  shell  and  easily  discernible  velum 
pigment).  Only  capsules  completely  full  of  eggs  and  containing  embryos  of  a  single 
age  class  were  included  in  experiments. 

Desiccation  tolerance  of  encapsulated  embryos  was  examined  at  two  relative 
humidities,  approximately  0%.  and  75%  as  determined  with  a  Honeywell  portable 
hygrometer.  These  relative  humidities  were  achieved  by  covering  the  bottoms  of 
glass  jars  with  anhydrous  CaSO4  or  a  saturated  solution  of  NaCl  in  distilled  water 
(O'Brien,  1948),  respectively.  Age-sorted  N.  obsoletus  capsules  were  spooned  into 
perforated,  plastic  petri  dishes,  and  most  of  the  adhering  water  quickly  blotted  away. 
From  50  to  209  egg  capsules  were  spooned  into  each  dish.  Large  clumps  of  cap- 
sules were  broken  up  into  smaller  groups,  but  no  attempt  was  made  to  isolate  each 
individual  capsule.  Each  dish  was  suspended  above  desiccant  in  a  jar  which  was 
then  sealed.  All  experiments  were  conducted  at  room  temperature,  22-23°  C. 
The  air  in  each  jar  was  stirred  by  rocking  the  dish  of  capsules  every  fifteen  minutes. 
At  pre-determined  intervals,  one  dish  from  each  age  class  was  removed  from  a  jar, 
capped  with  a  perforated  top,  and  submerged  in  running  sea  water  (approximately 
30/£e).  One  dish  of  egg  capsules  from  each  age  class  served  as  a  control  for  its 
age  group  and  was  not  subjected  to  desiccation,  being  placed  in  flowing  sea  water 
at  the  start  of  the  experiment.  Capsules  suspended  over  distilled  water  for  the 
entire  exposure  period  served  as  additional  controls;  relative  humidity  within  these 
jars  was  100%,  as  determined  with  a  Honeywell  portable  hygrometer.  After  treat- 
ment, the  dishes  of  egg  capsules  were  submerged  in  flowing  sea  water  and  examined 
periodically  to  assess  survival.  Tolerance  of  the  stress  was  indicated  by  the  eventual 
escape  of  veligers  from  the  capsules.  Capsules  do  not  open  spontaneously ;  a 
specialized  hatching  substance  produced  by  the  embryos  is  required  for  escape  to 
occur  (Pechenik,  1975). 

Additional  experiments  were  conducted  to  determine  the  effect  of  repeated  ex- 
posures to  low  humidity  air  on  pre-hatching  mortality.  Groups  of  capsules  were 


284  JAN  A.  PECHENIK 

subjected  to  75%  relative  humidity  lor  either  0.5  lir  or  2  hr  each  day,  until  hatching 
was  completed  after  about  ()  days. 

The  rates  of  water  loss  from  intertidal  A  .  ohsolctiis  egg  capsules  were  compared 
\vith  rates  of  water  loss  from  the  morphologically  similar,  but  primarily  subtidally- 
deposited  capsules  of  N.  trh'itlalits  (Scheltema  and  Scheltema,  1964).  The  mag- 
nitude of  the  differences  found  should  reflect  the  degree  to  which  N.  obsoletns 
capsules  are  specifically  adapted  for  their  intertidal  deposition.  Ten  to  12  min 
after  the  addition  of  CaSO4  desiccant  to  the  weighing  chamber  of  a  Cahn  Electro- 
balance,  a  single  egg  capsule  was  blotted  dry,  dropped  onto  the  weighing  pan,  and 
the  weighing  chamber  was  quickly  resealed.  Capsule  weight  was  determined  at 
30  sec  intervals  for  15  min,  or  until  the  weight  stopped  changing  between  readings. 

Data  were  obtained  for  40  egg  capsules,  ten  capsules  for  each  of  the  two  age 
groups  for  both  species.  The  rate  of  weight  loss  was  found  to  be  constant  for  all 
capsules  for  the  first  8.0  min  after  the  start  of  each  experiment.  Hence,  the  rate  of 
weight  loss  was  computed  from  the  weight  change  observed  during  this  interval. 
Capsule  weight  at  1.5  min  after  the  initiation  of  the  experiment  was  taken  as  the 
"initial"  weight.  Since  the  rate  at  which  water  can  be  lost  from  a  closed  container 
is  a  function  of  exposed  surface  area,  proportional  to  (weight)273,  the  data  were 
adjusted  using  the  following  expression  before  statistical  comparisons  were  made 
between  groups  of  capsules:  adjusted  rate  --  (mg  lost/30  sec)/(initial  capsule 
weight)273.  This  manipulation  substantially  reduced  variability  in  the  data.  Rates 
of  weight  loss  discussed  in  the  text  are  unadjusted  rates,  unless  otherwise  indicated. 

The  placement  of  N.  obsoletns  egg  capsules  with  respect  to  substrate  orientation 
was  examined  in  the  laboratory.  Adults  \vere  collected  from  Quissett  Harbor  and 
Barnstable  Harbor,  Massachusetts,  and  held  in  aerated  aquaria.  Snails  were  fed 
every  two  to  three  clays  on  chopped  Alcrccnaria  uicrccnaria  tissue.  A  rectangular, 
plastic  container  with  one  curled  edge  on  the  upper  surface  was  placed  on  the  bot- 
tom of  each  tank,  completely  submerged  in  sea  water,  and  the  number  of  capsules 
deposited  upon  each  surface  of  the  container  was  determined  on  nine  occasions  over 
a  2-week  period.  All  capsules  were  removed  from  the  container  after  each  observa- 
tion, so  that  the  extent  of  deposition  on  any  surface  was  never  limited  by  the  area 
available. 

The  extent  to  which  egg  capsules  are  protected  from  exposure  to  low  humidity 
air  in  the  field  was  estimated  by  spraying  exposed  clumps  of  Fucus  with  blue 
enamel  paint  during  low  tide  at  Quissett  Harbor  and  Eel  Pond.  The  spray  paint 
should  have  reached  only  those  capsules  exposed  to  the  air,  while  capsules  in  more 
sheltered,  high-humidity  sections  of  the  Fucus  should  have  remained  untouched  by 
the  paint.  The  Fucus  clumps  were  then  detached  from  their  rocks  and  the  numbers 
of  blue,  "speckled"  (exhibiting  from  one  to  several  small  blue  spots),  and  unpainted 
capsules  on  the  algae  wrere  determined.  The  position  of  each  capsule  with  respect 
to  the  basal  2  cm  of  algae  was  also  recorded. 

RESULTS 
Desiccation  tolerance  of  encapsulated  embryos 

Egg  capsules  of  N.  obsolctns  did  not  afford  substantial  protection  against  desic- 
cation. At  0%  relative  humidity,  mortalities  of  at  least  60%  were  sustained  after 


NASSARIUS  INTERTIDAL  DEVELOPMENT 


285 


TABLE  I 

Mortality  of  encapsulated  early-stage  N.  obsoletus  embryos  after  single  exposures  to  75%  relative 
humidity. 


Treatment 

Number  of  capsules 

Exposure  time  (hr) 

Mortality  (%) 

Sea  water  control 

209 

0 

0.5 

100%  relative  humidity  control 

111 

5 

0.0 

75%  relative  humidity 

156 

1 

0.0 

75%  relative  humidity 

123 

2 

3.3 

75%  relative  humidity 

126 

2.5 

1.6 

75%  relative  humidity 

151 

3 

2.6 

75%  relative  humidity 

141 

3.5 

12.5 

75%  relative  humidity 

167 

4.5 

10.2 

75%  relative  humidity 

169 

5 

15.7 

a  single  exposure  of  less  than  one  hour  (Fig.  1).  Suhstantial  mortality  of  early- 
stage  embryos  resulted  after  only  15  minutes  of  exposure.  Controls  exhibited  less 
than  5%  failure  to  hatch.  Even  at  75%  relative  humidity,  12%  of  the  capsules 
failed  to  release  larvae  after  a  single  3.5  hr  exposure,  while  control  mortalities  were 
less  than  0.5%  (Table  I).  Clumping  together  of  the  egg  capsules  in  the  petri 
dishes  probably  accounts  for  the  lack  of  increased  mortality  between  2  and  3  hours, 
since  clumps  would  retain  moisture  longer  than  individual  capsules.  Although 
mortalities  did  not  exceed  3. 3%  at  75%  relative  humidity  for  single  exposures  of 
less  than  3.5  hr,  daily  exposures  for  considerably  shorter  periods,  0.5-2  hr,  re- 
sulted in  mortalities  of  approximately  15%  (Table  II). 

Advanced-stage  encapsulated  embryos  were  significantly  more  tolerant  of  desic- 
cation stress  than  were  early-stage  embryos  (Fig.  1 ;  P  <  0.05,  as  tested  by  Chi- 
square).  This  indicates  that  age-related  changes  occur  either  in  the  water-re- 
taining ability  of  the  capsules  or  in  the  tolerance  of  the  embryos  themselves,  as  con- 
sidered below. 

Rates  of  water  loss  from  egg  capsules 

There  were  no  statistically  significant  differences  in  adjusted  rates  of  water  loss 
from  capsules  of  different  ages,  for  either  A/T.  obsoletus  or  AT.  trivittatus,  as  analyzed 

TABLE  II 

Mortality  of  encapsulated  N.  obsoletus  after  daily  exposures  to  75%  relative  humidity.  Capsules 
contained  early-stage  embryos  at  start  of  experiment. 


Treatment 

Number  of  capsules 

Exposure  time  (hr) 

Mortality  (%) 

Control 

72 

0 

4.2 

Control 

131 

0 

3.0 

100%  relative  humidity 

122 

1 

4.1 

100%  relative  humidity 

50 

2 

0.0 

100%  relative  humidity 

143 

2 

0.0 

75%  relative  humidity 

63 

0.5 

17.5 

75%  relative  humidity 

73 

2 

12.3 

286 


JAN  A.  PECHEN1K 


100  r 


tr 
o 


—  ?             1 

i               i               i 

D 

9 

15               30 

45                60               75 

90 

MINUTES 

DESICCATED 

FIGURE  1.  Desiccation  tolerances  of  encapsulated  embryos  of  N.  obsoletus.  Experiments 
were  conducted  over  CaSOi  desiccant.  Circles  represent  data  from  capsules  containing  late- 
stage  embryos  (N  =  455  capsules).  Squares  represent  data  from  capsules  containing  early- 
stage  embryos  (N  =  293  capsules).  Open  symbols  indicate  data  from  control  capsules  beld 
at  100%  relative  humidity  for  the  full  90  minutes. 


by  analysis  of  variance  (FA-.  tnnttatus  -'-  1.94;  FjV  0^.S.(,;,^/(S  —  2.62;  P  >  0.1  ;  N  :-  20 
capsules  for  each  species).  Since  the  water-retaining  ability  of  the  capsule  itself 
does  not  change  with  age.  age-related  improvements  in  desiccation  tolerance  are 
attributable  to  a  change  in  embryonic  tolerance,  possibly  due  to  the  development 
of  the  embryonic  shell. 

Egg  capsules  of  both  species  lost  weight  at  a  constant  rate  for  at  least  the  first 
8  minutes  of  observation.  One-way  analysis  of  variance  revealed  no  statistically 
significant  differences  in  the  adjusted  rates  of  water  loss  from  egg  capsules  of  the 
two  different  species  (F  =  =1.07;  N  —  20  capsules  for  each  species;  P>0.1). 
Thus,  the  water-retaining  ability  of  the  intertidally-placed  N.  obsoletus  capsule  is 
essentially  identical  to  that  of  the  subtidally-placed  N.  trlvittatus  capsule. 

Rates  of  water  loss  (weight/unit  time)  from  capsules  similar  in  initial  weight 
were  essentially  identical,  regardless  of  capsule  age  or  identity.  Larger  capsules 
lost  weight  more  rapidly  than  smaller  capsules,  as  expected.  The  relationship 
between  the  rate  of  weight  loss  (Y)  and  the  weight  of  the  capsule  at  1.5  minutes 
after  the  start  of  the  experiment  (X)  is  given  by  the  equation  Y  =  0.027  +  0.01 5X, 
as  calculated  by  linear  regression  analysis  (Fig.  2).  The  mean  rate  of  weight  loss 
from  the  40  egg  capsules  was  0.044  mg/30  sec  ±  0.005  (mean  ±  s.d.).  The  dif- 
ferences in  rates  of  water  loss  between  egg  capsules  were  due  primarily  to  differ- 
ences in  the  sizes  of  the  capsules;  the  correlation  coefficient  (r)  between  rate  of 


NASSARIUS  INTERTIDAL  DEVELOPMENT 


287 


weight  loss   (mg/30  sec)   and  "initial"  egg  capsule  weight  was  0.82   (F  =  80.85; 
N  -  •  40  experiments). 

Placement  of  egg  capsules  in  (lie  laboratory 

Capsules  were  not  deposited  randomly  on  the  plastic  containers  (Table  III). 
The  differences  in  the  numbers  of  capsules  received  by  each  surface  of  the  container 
are  significant  at  the  0.01  level,  as  tested  by  analysis  of  variance  (d.f.  =  =  11,96;  F  = 
4.7).  Statistical  comparisons  reveal  two  behavioral  phenomena  associated  with 
capsule  deposition.  Low  thigmokinesis  (Fraenkel  and  Gunn,  1961)  is  revealed  by 
particularly  heavy  deposition  along  edges,  under  the  curled  edge,  or  underneath 
the  platform,  regions  maximizing  the  amount  of  contact  stimulation  of  the  foot  of 
the  depositing  female.  There  is  also  an  orientation  component,  specifically  a  prefer- 
ence for  depositing  while  hanging.  The  fewest  capsules  were  attached  to  the  bottom 
and  outside  top  surfaces  of  the  container.  The  correlation  coefficient  (r)  between 
the  number  of  capsules  deposited  and  available  surface  area  on  the  different  surfaces 
of  the  container  was  0.36.  indicating  that  placement  preferences  were  not  related  to 
the  amount  of  surface  area  available. 

Capsules  were  never  deposited  above  the  water  line  in  laboratory  aquaria,  sug- 
gesting that  capsule  deposition  in  the  field  occurs  only  when  the  substrate  is  sub- 
merged. 


u 

o 
(0 
O 
to 


CO 
O 


X 
<£ 

UJ 


UJ 


0.060 


0.055 


0.050 


0.045 


0.040 


0.035 


0.030 


0.6 


D 


O 


O 


0.8 


1.2 


1.4 


1.6 


1.8 


2.0 


2.2 


CAPSULE     WEIGHT     (mg) 


FIGURE  2.  Rate  of  weight  loss  for  desiccating  egg  capsules  of  Ar.  obsolctus  (squares)  and 
N.  triz'ittatus  (circles)  as  a  function  of  "initial"  capsule  weight.  Open  symbols  represent  data 
obtained  from  capsules  containing  early-stage  emryos,  while  solid  symbols  represent  data  ob- 
tained from  capsules  containing  advanced-stage  embryos. 


288  JAN  A.  PECHEN1K 

TABLE  III 

Distribution  of  N.  obsolctus  egg  capsules  deposited  on  plastic  containers  in  the  laboratory. 


Quissett  Harbor 
adults 

Barnstable  Harbor 
adults 

Total  capsules  deposited 

855 

501* 

Percentage  of  capusles  deposited  on: 

Bottom 

1.6 

2.0 

Back 

18.5 

4.7 

Sides 

9.8 

6.1 

Top  (inside) 

16.0 

17.2 

Top  (outside) 

1.2 

0.0 

Edges  (inside) 

21.9 

64.0 

Edges  (outside) 

11.0 

5.0 

Under  curled  edge 

20.0 

1.0 

This  container  was  improperly  anchored.  An  additional  101  capsules  were  deposited  on  the 
underside  of  the  container. 

Placement  of  capsules  in  tJie  field 

Due  to  the  low  thigmokinetic  component  of  the  adult  egg-laying  behavior,  one 
expects  to  find  more  capsules  deposited  near  the  holdfast  of  FIICHS  than  upon  other 
portions  of  the  algae,  owing  to  the  closeness  of  individual  strands  in  the  holdfast 
area  relative  to  other  sections  of  the  algae  when  submerged.  One  would  also 
expect  most  of  the  capsules  to  be  deposited  on  the  undersides  of  the  strands,  due  to 
the  apparent  preference  for  hanging  while  depositing. 

Capsules  on  Fncus  collected  at  Quissett  Harbor  were  located  primarily  in  the 
region  of  the  holdfast,  as  predicted  (Table  IV).  Although  this  did  not  hold  true 
for  Eel  Pond  FIICIIS  on  a  strictly  numerical  basis,  many  more  capsules  were  de- 
posited in  the  holdfast  region  than  elsewhere  on  the  algae  when  the  relative  surface 
areas  (estimated  by  dry  weight)  available  for  deposition  of  capsules  were  con- 
sidered. One  hundred  capsules/gram  algae  were  found  in  the  region  of  the  holdfast 
and  11.6  capsules/gram  algae  were  found  elsewhere,  based  on  examination  of  four 
plants,  and  counts  of  68  and  235  capsules  in  the  basal  and  distal  regions  of  the  algae, 
respectively. 

The  pattern  of  paint  on  egg  capsules  was  similar  for  tufts  of  FIICIIS  sprayed  at 
both  locations  at  low  tide,  so  that  only  the  results  from  Quissett  Harbor  are  pre- 
sented here  (Table  IV).  Most  of  the  capsules  were  at  least  partially  protected 
from  exposure  to  low  humidity  air,  fewer  than  8%  of  the  capsules  being  entirely 

TABLE  IV 

Distribution  and  color  of  583  N.  obsoletus  egg  capsules  recovered  from  Fucus  spray-painted  at  low  tide 
at  Quissett  Harbor,  Massachusetts. 

Holdfast         Nonholdfast 


Deposited  in  zone 

69.0% 

31.0% 

Blue  capsules 

7.7% 

4.4% 

Unpainted  capsules 

8.0% 

13.3% 

Speckled  capsules 

84.3% 

82.3% 

NASSARIUS  INTERTIDAL  DEVELOPMENT  289 

exposed  at  low  tide.     Speckled  capsules,  which  generally  exhibited  only  one  or  two 
small  spots  of  paint,  can  probably  be  considered  as  being  protected. 

The  dearth  of  firm  substrate  available  to  the  large  N.  obsolctus  population  on 
the  Barnstable  mudflats  results  in  the  attachment  of  capsules  to  all  available  surfaces; 
suitable  macro-algae  are  literally  covered  with  capsules.  Since  much  of  this  ma- 
terial is  completely  exposed  to  air  at  low  tide,  pre-hatching  mortality  here  may  be 
high.  Dried  and  shrunken  A7,  obsolctus  egg  capsules  are  frequently  encountered. 
Of  664  egg  capsules  collected  at  the  end  of  a  low  tide  at  Barnstable  early  in  August 
and  held  in  flowing  sea  water  in  the  laboratory,  28%  failed  to  release  veligers.  As 
most  of  these  capsules  contained  early-stage  embryos  when  collected  and  because 
repeated  exposure  to  low  humidity  air  increases  mortality  significantly,  as  demon- 
strated above,  28%  is  probably  a  minimum  estimate  of  the  total  pre-hatching  mor- 
tality that  would  have  occurred  had  the  embryos  been  allowed  to  complete  their 
development  in  the  field. 

DISCUSSION 

Caution  must  be  used  in  relating  experimental  data  on  embryonic  desiccation 
tolerances  and  rates  of  water  loss  from  egg  capsules  to  actual  events  in  the  field. 
Relative  humidities  of  0%  are  unlikely  to  occur  normally,  although  relative  humidi- 
ties of  40%  are  not  uncommon  in  the  intertidal  zone  (Kanwisher.  1957).  Experi- 
ments conducted  at  0%  relative  humidity  demonstrate  two  important  points,  how- 
ever. First,  encapsulated  embryos  of  N.  obsolctus  are  more  susceptible  to  desicca- 
tion stress  than  are  the  embryos  of  Mclani^us  bidcntatits  (Holle  and  Dineen,  1957). 
The  apparently  lower  susceptibility  of  M.  bidcntatits  embryos  to  desiccation  is 
surprising;  although  the  egg  masses  are  deposited  high  in  the  intertidal  zone,  an 
accumulation  of  detritus  around  them  probably  prevents  their  desiccation  (Russell- 
Hunter,  Apley  and  Hunter,  1972).  Secondly,  the  egg  capsules  of  A7.  obsolctus  are 
no  more  effective  in  retaining  water  than  are  capsules  of  the  subtidal  N.  trh'ittatus. 
Bayne  (1968)  obtained  similar  results  with  the  egg  capsules  of  the  intertidal  gastro- 
pod Nucclla  lapillus.  Rates  of  water  loss  [mg/(min-unit  surface  area)]  from  egg 
capsules  of  this  species  were  nearly  identical  with  those  from  the  normally  sub- 
merged spawn  of  both  the  opisthobranch  Aplysia  punctata  and  the  basommatophoran 
Lyninca  staynalis. 

It  is  not  possible  to  predict  the  actual  extent  of  pre-hatching  mortality  from  the 
experimental  data  presented  here.  The  impact  of  desiccation  in  the  field  will  vary 
with  humidity,  temperature,  duration  and  degree  of  exposure,  wind  velocity,  and  the 
thickness  of  any  boundary  layer  that  may  be  present  above  the  capsules.  The  im- 
portant points  are  that  substantial  pre-hatching  mortality  of  A",  obsolctus  occurs  in 
the  laboratory  after  even  a  single,  short  exposure  to  75  7  relative  humidity,  and 
that  daily  exposure  to  desiccation  results  in  significantly  greater  mortality  than  that 
observed  after  a  single  exposure.  Substrates  literally  encrusted  with  N.  obsoletus 
egg  capsules  are  often  found  completely  out  of  the  water  at  low  tide  on  the  Barn- 
stable  mudflats.  In  such  cases,  pre-hatching  mortality  must  be  high,  since  the  egg 
capsules  are  not  effective  in  preventing  water  loss  and  the  embryos  are  not  particu- 
larly tolerant  of  dehydration.  Spight  (1975)  reported  pre-hatching  mortalities  of 


2W)  JAN  A.  PECHENIK 

approximately  40%  for  intertidally-deposited  Thais  laincllosa  egg  capsules;  desicca- 
tion was  a  major  cause  of  this  mortality. 

There  is,  therefore,  no  evidence  that  gastropod  egg  capsules  are  specifically 
adapted  for  placement  in  the  intertidal  zone,  and  it  seems  unlikely  that  egg  capsules 
evolved  as  adaptations  to  intertidal  stresses. 

Protection  of  the  encapsulated  embryos  of  N.  obsolctns  seems  dependent  upon 
adult  behavior.  Because  capsules  are  placed  on  the  undersides  of  FHCIIS,  they  are 
kept  moist  by  the  blanketing  effect  of  the  seaweed  above  them.  Detailed  experi- 
mental work  on  egg  capsule  placement  preferences  is  lacking  for  most  marine  in- 
vertebrates (Meadows  and  Campbell,  1972),  but  the  protection  of  early  develop- 
mental stages  through  adult  spawning  behavior  appears  widespread  in  marine 
gastropods.  Kudinsky  (1972)  claims  that  eggs  of  the  prosobranch  gastropod 
Testudinalia  tcsscllata  are  laid  preferentially  in  situations  where  they  are  spared 
direct  exposure  to  sunlight.  The  egg  capsules  of  Conns  >?/>/>.  (Kohn,  1961), 
Cypraea  spp.  (Crovo,  1971),  shallow-water  columbellid  gastropods  (Bandel,  1973), 
Urosalpinx  cincrca  (Carriker,  1955),  and  Bciiibiciiun  uiinitiiin  (Anderson,  1962) 
are  deposited  on  the  undersides  of  rocks,  presumably  to  prevent  desiccation  (Ander- 
son, 1962;  Kohn,  1961),  and  the  archaeogastropod  Ncritina  virginea  is  said  to  de- 
posit its  capsules  preferentially  in  crevices  (Andrews,  1935). 


This  work  was  supported  by  a  graduate  fellowship  from  the  Woods  Hole 
Oceanographic  Institution.  Laboratory  space  was  provided  by  Dr.  R.  Scheltema. 
The  manuscript  has  benefitted  greatly  from  the  helpful  comments  of  L.  Halderman, 
R.  Harbison,  R.  Hoffman,  D.  Miller,  T.  Murray,  F.  Perron,  D.  Pratt,  W.  D. 
Russell-Hunter,  R.  Scheltema,  and  an  anonymous  reviewer. 

SUMMARY 

1.  The  extent  to  which   reproduction  of  the  intertidal   mud   snail,   Nassarius 
obsoletus,  is  adapted  to  the  intertidal  environment  was  examined  in  an  attempt 
to  understand  the  adaptive  significance  of  egg  capsules  in  the  life  history. 

2.  Contrary  to  expectation,  laboratory  studies  on  desiccation  tolerance  of  en- 
capsulated embryos  and  rates  of  water  loss  from  egg  capsules  failed  to  reveal  any 
adaptation  to  intertidal  development.     Fifteen  minutes  of  desiccation  over  CaSC>4 
caused  as  much  as  20%  mortality  of  N.  obsolctns  embryos,  and  daily  0.5  hr  ex- 
posures to  75%  relative  humidity  killed  17.5%  of  the  embryos.     Egg  capsules  of 
N.  obsoletus  and  those  of  the  subtidal  N.  trivittatits  lost  water  at  essentially  equal 
rates. 

3.  Protection  of  the  developing  embryos  seems  dependent  upon  adult  behavior. 
Adults  tend  to  deposit  egg  capsules  into  microenvironments  where  the  embryos 
are  probably  spared  exposure  to  desiccation  stress  at  low  tide.     Fewer  than  8%  of 
the  capsules  examined  at  Ouissett  Harbor,  Massachusetts,  were  fully  exposed  to 
desiccation. 

LITERATURE  CITED 

AIYAR,  R.  G.,  1931.     An  account  of  the  development  and  breeding  habits  of  a  brackish  water 
polychaete  worm  of  the  genus  Marfhyxa.  J.  Linn.  Soc.  Loud.  Zool.,  37:  387-403. 


NASSARIUS  INTERTIDAL  DEVELOPMENT  291 

ANDERSON,  D.  T.,  1962.    The  reproduction  and  early  life  histories  of  the  gastropods  Bembicium 

auratitin    (Quoy   and    Gaimard)     (Fain.    Littorinidae),    Cellana    tramuscrica    (Sower) 

(Fam.    Patellidae)    and   Melanerita    mclaimtragus    (Smith)     (Fam.    Neritidae).     Proc. 

Linn.  Sac.  N.S.I  I '..87:  62-68. 

ANDREWS,  E.  A.,  1935.    The  egg  capsules  of  certain  Neritidae.    /.  M  or  filial.,  57:  31-59. 
BANDEI.,  K.,  1973.     Spawning  and  development  of  some  Columbellidae  from  the  Caribbean  Sea 

of  Columbia  (South  America) .     V  diner.  16:    271-282. 
BAYNE,   C.  J.,   1968.     A   study  of  the  desiccation  of  egg  capsules   of  eight  gastropod   species. 

J.Zool.Lomi.,  155:401-411. 
CARRIKER,   M.   R.,    1955.      Critical   review   of  biology  and  control   of  oyster  drills,    Urosalfiin.v 

einerca  and  Enfileura  caudata.    U.  S.  Fish  JVildl.  Serv.  Spec.  Sci.  Rep.,  148:  1-150. 
COSTELLO,  D.  P.,  AND  C.  HENLEY,  1971.     Methods  for  obtaining  and  handling  marine  eggs  and 

embryos.  Second  Ed.    Baker  Manufacturing  Company,  Massachusetts,  247  pp. 
CROVO,    M.   E..    1971.      C\praca    cerrus   and   Cyfiraca    zebra    in    Florida — one    species    or    two? 

Vdigcr.  13 :  292-295. 

DIMON,  A.  C.,  1905.     The  mud  snails:    Xassa  obsoleta.    Cold  Spring  Harbor  Monogr.,  5:  4-51. 
FRAENKEL,  G.  S.,  AND  D.  L.  GUNN,  1961.     The  orientation  of  animals — kineses,  taxes  and  com- 
pass reactions.    Dover  Publications,  New  York,  376  pp. 
FRETTER,  V.,  AND  A.  GRAHAM,   1962.     British  prosobrancli   mull  uses — their  functional  anatomy 

and  ecology.    Adlard  and  Son,  Ltd.,  Bartholomew  Press,  Dorking,  England,  755  pp. 
GALTSOFF,  P.  S.,  H.  F.  PRYTHERCH,  AND  J.  B.  ENGLE,  1937.     Natural  history  and  methods  of 

controlling  the   common   oyster   drills    Urosalpin.r  cinerca    Say   and   Euplcura   caudata 

(Say).     U.  S.  Fish.  Wildl.  SOT.  Circ..  25:  1-24. 
GIBBS,  P.  E.,  1968.    Observations  on  the  population  of  Scoloplos  armiqcr  at  Whitsable.    /.  Mar. 

Biol.  Assoc.  U.K.,  48  :  225-254. 
HOLLE,  P.  A.,  AND  C.  F.  DINEEN,  1957.     History  of  the  salt  marsh  snail  Melampns  bidcntatus. 

Nautilus,  70 :  90-95. 

HOUBRICK,  R.  S.,  1973.     Studies  on  the  reproductive  biology  of  the  genus  Ccrithium    (Gastro- 
poda: Prosobranchia)   in  the  western  Atlantic.     Bull.  Mar.  Sci.,  23:  875-904. 
HUNT,   S.,   1966.     Carbohydrate  and  amino  acid  composition  of  the  egg  capsule  of  the  whelk 

Buccinum  undatum  L.  Nature,  210:  436-437. 

KANWISHER,  J.,  1957.     Freezing  and  drying  in  intertidal  algae.     Biol.  Bull.,  133:  275-285. 
KOHN,  A.  J.,  1961.     Studies  on  spawning  behavior,  egg  masses,  and  larval  development  in  the 

gastropod  genus   Conns.   I.  Observations  on  nine  species  in  Hawaii.    Pacific  Sci.,  15: 

163-180. 
KUDINSKY,    O.    Y.,    1972.      Reproduction   and    gametogenesis    of    Testudinalia    tessellata    Mosk 

(Prosobranchia,  Docoglossa)   from  a  coastal  Barents  Sea  Province.    Dokl.  Biol.  Sci., 

202  :  57-60. 
MEADOWS,  P.  S.,  AXD  J.  I.  CAMPBELL,   1972.     Habitat  selection  by  aquatic  invertebrates.    Adv. 

Mar.  Bio!..  10:  271-382. 
AfiLEiKOVSKY,  S.  A.,  1971.     Types  of  larval  development  in  marine  bottom  invertebrates,  their 

distribution  and  ecological  significance:    a  reevaluation.    Mar.  Biol.,  10:  193-213. 
O'BRIEN,  F.  E.  M.,  1948.    The  control  of  humidity  by  saturated  salt  solutions.    /.  Sci.  lustrum., 

25  :  73-76. 
PECHENIK,  J.  A.,  1975.     The  escape  of  veligers  from  the  egg  capsules  of  Nassarius  obsolctus 

and  Nassarius  trivittatns   (Gastropoda,   Prosobranchia).     Biol.  Bull.,  149:   580-589. 
RUSSELL-HUNTER,   W.   D.,   M.    L.   APLEY,   AND   R.    D.   HUNTER,    1972.     Early   life-history   of 

Mchiinpns  and  the  significance  of  semilunar  synchrony.    Biol.  Bull.,  143:  623-656. 
SCHELTEMA,  R.  S.,  1962.     Pelagic  larvae  of  New  England  intertidal  gastropods.     I.  Nassarius 

obsolctus  Say  and  Nassarius  1'ibc.r  Say.  Trans.  Am.  Microsc.  Soc.,  81:   1-11. 
SCHELTEMA,  R.   S.,  AND  A.   H.   SCHELTEMA,   1964.     Pelagic  larvae  of  New  England   intertidal 

gastropods.    III.  Nassarius  tr'nittatus.    Hydrobiologia,  25:  321-329. 
SPIGHT,  T.  M.,   1975.     Factors  extending  gastropod  embryonic  development  and  their  selective 

cost.    Oecologia,  21 :  1-16. 
SVERDRUP,   H.   U.,   M.  W.  JOHNSON.   AND   R.   H.   FLEMING,    1942.     The  oceans:    their  physics, 

chemistry  and  general  biology.    Prentice-Hall,  New  Jersey,  1087  pp. 


KdVrence:   Blol  Bull.,  154:  292-301.    (April,  1978) 


DIFFUSIONAL  WATER  PERMEABILITY  IN  SELECTED 

MARINE  BIVALVES 

ROHKRT  1).  PRUSCH  AND  CAROL  HALL 

Division  of  Bioloi/y  mid  Medicine,  Brcmm  University,  Providence,  Rhode  Island  02912; 
and  the  Marine  Biological  Laboratory,  Woods  Hole  Massachusetts  02543 

The  intertidnl  habitat  presents  many  severe  problems  to  the  organisms  living 
in  this  particular  environment.  Cyclic  fluctuations  in  salinity,  oxygen  levels  and 
temperature  occur,  and  these  changes  in  the  physical  parameters  of  this  environ- 
ment are  amplified  with  increasing  height  above  the  sublittoral  zone.  Organisms 
which  inhabit  the  intertidal  area  have  undergone  certain  adaptations  which  in- 
crease their  survival  potential  in  the  face  of  a  constantly  changing  environment. 

In  crustaceans,  the  permeability  of  the  body  surface  to  water  and  ions  can 
be  correlated  with  the  animals'  particular  habitat  (Lockwood,  1962;  Herreid, 
1969a,  b).  In  this  case,  sublittoral  crustaceans  are  more  permeable  than  littoral 
species,  which  in  turn  are  more  permeable  than  estuarine  species.  Pieces  of  exo- 
skeleton  from  crustaceans  which  are  osmoregulators,  and  for  the  most  part  inter- 
tidal,  are  less  permeable  than  osmoconformers,  which  are  generally  sublittoral 
(Gross,  1957).  In  barnacles,  the  resistance  of  the  organism  to  desiccation  is  a 
function  of  its  vertical  zonation  level,  i.e.,  the  higher  the  species  in  the  intertidal 
zone,  the  more  resistant  it  is  to  desiccation  (Foster,  1971).  Some  intertidal 
crustaceans  apparently  are  capable  of  controlling  the  permeability  of  the  exoskele- 
ton  and  decrease  water  permeability  in  response  to  decreased  external  salinity 
(Smith,  1970;  Lockwood,  Inman  and  Gmrtenay,  1973). 

Physiological  adaptations  to  some  environmental  stress  situations  have  also  been 
demonstrated  in  intertidal  molluscs.  Resistance  to  desiccation  in  certain  intertidal 
gastropods  has  been  correlated  with  vertical  zonation  (Brown,  1960),  and  the  se- 
quence of  thermal  death  points  in  some  intertidal  gastropods  has  also  been  correlated 
with  zonation  level  (Broekhuysen,  1940).  Physiological  adaptations  at  the  tissue 
level  have  also  been  demonstrated  in  intertidal  bivalve  molluscs  (Vernberg, 
Schlieper  and  Schneider,  1963).  In  this  case  bivalve  gill  tissue  from  intertidal 
animals  could  withstand  a  wider  range  of  salinity  changes  than  could  sublittoral 
animals  as  determined  by  gill  ciliary  activity.  Behavioral  adaptations  of  intertidal 
bivalves  include  closing  of  the  valves  during  periods  of  osmotic  stress  (Krogh, 
1939). 

A  question  arises  as  to  whether  or  not  the  tissues  of  intertidal  bivalve  molluscs 
display  the  same  adaptive  water  permeability  characteristics  as  intertidal  crustaceans, 
i.e.,  a  decrease  in  water  permeability  with  increasing  height  in  the  intertidal  zone; 
or  whether  the  distribution  of  intertidal  bivalves  is  influenced  mainly  by  other  en- 
vironmental factors  independent  of  salinity  (Pilgrim,  1953).  In  addition,  can 
specific  intertidal  bivalves  alter  the  diffusional  water  permeability  of  their  tissues 
in  response  to  an  osmotic  stress,  as  is  apparently  the  case  in  crustaceans?  These 

292 


BIVALVE  WATER  PERMEABILITY  293 

problems  were  investigated  in  a  series  of  marine  bivalves  by  measuring  the  rate  of 
movement  of  tritiated  water  across  the  isolated  mantle  tissue. 

MATERIALS  AND  METHODS 

The  diffusional  water  permeability  across  isolated  mantle  tissue  was  determined 
in  eight  different  lamellibranch  species  obtained  from  different  collection  sites  in 
New  England.  Water  samples  were  also  collected  from  each  collection  site  and 
total  salinity  determined  by  potentiometric  chloride  titration.  Specimens  of  Placopec- 
tcn  magellenicus  (Gmelin)  and  Modiohts  niodioltts  (L.)  were  both  collected  from 
waters  off  Mount  Desert  Island,  Maine.  Placopcctcn  was  collected  from  a  depth  of 
10  m,  while  M.  inodiolits  was  found  at  a  depth  of  2  to  3  m,  in  both  cases  the 
salinity  of  the  water  was  approximately  32(/cr.  Specimens  of  Spisnla  solidissiina 
(Dillwyn)  were  collected  by  the  Supply  Department  of  the  Marine  Biological 
Laboratory,  Woods  Hole,  Massachusetts.  Spisnla  was  found  approximately  2  m 
deep  on  a  sandy  bottom  in  water  with  an  average  salinity  of  33/£<?.  This  population 
of  Spisnla  was  never  exposed  at  low  tide.  Specimens  of  Mercenaria  incrccnaria  (L.) 
were  collected  from  Narragansett  Bay,  Rhode  Island,  in  water  0.5  to  2  m  deep 
with  muddy-sand  substratum  and  having  a  salinity  range  from  24  to  31//c.  Two 
different  groups  of  Mytilits  cdulis  (L.)  were  collected  from  the  rocky  shores  off 
Jamestown,  Rhode  Island.  One  group  was  located  low  in  the  intertidal  zone 
(ELWN),  while  the  other  group  was  found  sublittorally  in  approximately  5  m 
water.  The  salinity  of  the  water  at  this  site  ranges  from  26  to  32C/(C.  Specimens 
of  Modiolus  deinissus  (Dillwyn),  Crassosfrca  virginica  (Gmelin)  and  Mya  arenaria 
(L.)  were  all  collected  from  Sippiwissett  salt  marsh  near  Woods  Hole  and  speci- 
mens of  Anodonta  sp.,  also  used  in  this  study,  were  obtained  from  a  local  biological 
supply  house.  Mantle  water  permeability,  unless  indicated  otherwise,  was  measured 
in  freshly  collected  animals. 

The  mantle  was  used  in  these  studies  because  of  its  relatively  simple  structure, 
an  epithelial  sheet  with  two  cell  layers  (Neff,  1972),  and  the  ease  with  which  it  can 
be  removed  from  the  animal.  The  mantle  was  excised  from  the  animal  by  cutting 
the  adductor  muscles  to  open  the  valves  and  then  removing  the  central  portion  of 
the  mantle  from  one  of  the  valves  where  it  was  not  attached.  The  mantles  were 
always  covered  with  sea  water  during  dissection  and  experimental  procedure,  ex- 
cept for  Anodonta  for  which  a  Ringer's  solution  was  used  (Istin  and  Kirschner, 
1968). 

The  isolated  mantle  was  used  to  separate  a  diffusion  chamber  into  two  com- 
partments of  10  ml  each,  the  diameter  of  the  exposed  tissue  being  0.5  cm.  The 
diffusional  water  permeability  (P,i)  of  the  isolated  mantle  was  determined  by  adding 
10  to  50  yu.Ci  tritiated  water  (THO)  to  one  compartment  and  monitoring  its  rate  of 
appearance  in  the  other  compartment.  Both  compartments  were  constantly  stirred 
by  bubbling  air  through  the  medium.  Aliquots  (100  /j.\}  of  the  medium  in  the 
second  compartment  were  taken  at  various  time  intervals,  placed  in  a  scintillation 
vial  with  15  ml  Aquasol  (New  England  Nuclear)  and  counted  in  a  liquid  scintilla- 
tion counter.  Each  aliquot  removed  during  the  course  of  an  experiment  was  re- 
placed with  an  equal  volume  of  fresh  medium  in  order  to  maintain  constant  volume. 


294 


R.  D.  PRUSCH  AND  C.  HALL 


Samples  of  the  medium  in  the  compartment  to  which  THO  was  initially  added 
were  also  counted.  Since  identical  solutions  are  on  both  sides  of  the  mantle  and  it 
is  assumed  that  the  concentration  of  THO  added  to  the  first  compartment  remains 
constant  for  the  duration  of  the  experiment,  P(1  can  he  calculated  from  the  following 
relationships:  first,  specific  activity  (SA)  --  [THOJ/fHaO]  ;  secondly,  tritiated 
water  flux,  JTIIO  ~  slope  of  the  linear  portion  of  the  flux  curve/mantle  surface  area ; 
thirdly,  water  flux,  Jn2o  - :  JTHO/SA  ;  and  fourthly,  diffusional  water  permeability, 
Pa  =:  Jn2o/[HoO].  Mantle  thickness  was  determined  by  freezing  isolated  pieces 
of  tissue  in  an  acetone-dry  ice  hath,  then  measuring  the  thickness  of  broken  pieces 
of  the  frozen  mantle  through  a  microscope  with  a  calibrated  ocular  micrometer. 

Specimens  of  Mytilus  cditlis,  collected  by  the  MBL  Supply  Department  from 
Lucas  Shoal  off  of  Martha's  Vineyard  at  a  depth  of  approximately  9  m,  were  used 
in  another  series  of  experiments  designed  to  determine  whether  or  not  this  par- 
ticular bivalve  species  is  capable  of  altering  its  tissue  water  permeability  in  response 
to  an  induced  environmental  osmotic  stress.  Approximately  fifty  individuals  were 
randomly  divided  into  two  groups  after  the  mantle  water  Pd  was  initially  determined 
for  five  individuals.  One  group  of  animals  was  maintained  in  ten  liters  of  aerated, 
full-strength  sea  water,  while  the  other  group  was  maintained  in  an  equal  volume  of 
70%  sea  water.  The  water  for  both  groups  of  animals  was  changed  daily.  Mantle 


6  r 


o 


e 

CL 
o 


20 


40 


60 


mm 


80 


100 


120 


FIGURE  1.  Unidirectional,  tritiated  water  movement  across  an  isolated  piece  of  Mercenaria 
mercenaria  mantle  tissue.  The  total  counts  per  minute  appearing  in  the  bath  are  presented  as 
a  function  of  time. 


BIVALVE  \YATER  PERMEABILITY 


water  Pa  was  determined  for  animals  from  both  groups  at   various  time  intervals 
after  the  start  of  the  equilibration  period. 

In  all  of  the  experiments  in  this  study,  no  selection  for  size  was  made  of  the 
animals  used,  except  that  each  animal  had  to  be  large  enough  to  yield  a  piece  of 
mantle  tissue  which  would  cover  the  hole  between  the  chambers.  All  experiments 
were  performed  at  room  temperature,  19  to  20°  C.  In  most  instances  this  tem- 
perature was  higher  than  the  environmental  temperature  the  animals  were  col- 
lected at,  but  most  likely  this  temperature  change  has  a  minimal  effect  on  the  mea- 
sured mantle  water  Pd.  For  example,  the  diffusion  of  water  between  10  and  20°  C 
is  not  greatly  influenced  by  an  increase  in  temperature,  Q10  =:  1.04.  The  results 
are  presented  as  the  mean  (number  of  determinations)  ±  the  standard  error  of  the 
mean. 

RESULTS 

A  representative  THO  flux  curve  is  shown  in  Figure  1.  THO  was  added  to 
one  side  of  the  flux  chamber  separated  into  two  compartments  by  the  isoated  mantle 
tissue,  and  its  rate  of  appearance  followed  on  the  other  side  of  the  chamber.  Since 
the  amount  of  isotope  added  to  the  first  compartment  is  relatively  large,  and  in  the 
time  course  of  these  experiments  it  decreases  only  slightly  (less  than  5%  ),  the  flux 
curve  is  linear  after  an  initial  lag  period.  The  observed  lag  period  in  these  experi- 
ments, 3  to  10  minutes,  is  most  likely  due  to  the  initial  time  required  for  a  constant 
specific  activity  to  be  established  in  the  intracellular  compartments  of  the  isolated 
mantle  tissue  and  did  not  vary  in  the  different  species  examined  in  this  study. 


o 

OJ 


E 
o 

10 

O 


10 
8 
6 
4 
2 


o 


o 


o 


o 


o 


o 


o 


M.  DEMISSUS 

CRASSOSTREA 


A 

MYA 


A 

SPISULA 


MERCENARIA  SPISULA  PLACOPECTEN 

MYTILUS  (HIGH)  M.  MODIOLUS  MYTILUS(LOW) 


INTERTIDAL- 


(EHWS)  (EHWN) 


;ELWN)    (ELWS) 


SUBLITTORAL 


FIGURE   2.      Calculated    diffusional    water   permeabilities    (P,i)    for    isolated    mantle    tissue 
from  a  series  of  lamellibranch  molluscs  as  a  function  of  their  relative  distribution  to  each  other. 


2%  K.   l>.    PRUSCH  AND  C.  HALL 

TABLE  I 

M 'untie  thit'knt'ss  in  a  scries  of  bivalves  in  which  the  diffusional  water  permeabilities  were  also  deter- 
mined. 

Mantle  thickness 
mm 

Modiolus  demissus  1.15  ±  0.06  (12) 

Myt ilns  edulis  (high)  1.28  db  0.09  (  6) 

Mercenariti  niercenaria  0.76  ±  0.05  (10) 

Modiolus  modiolus  1.04  ±0.07  (   7) 

Spisula  solidissima  0.73  ±  0.02  (  6) 

Anodonta  sp.  0.68  ±  0.05  (  7) 


Calculated  diffusional  water  permeabilities  for  eight  different  marine  lamelli- 
branch  species  are  presented  in  Figure  2  as  a  function  of  their  distribution  in  rela- 
tion to  one  another.  These  Pd  values  are  as  follows:  Modiolus  dcuiissus,  2.2  ± 
0.18  (12)  ;  Crassostrca  virginica,  3.01  ±  0.28  (7)  ;  Mya  arenaria,  5.51  +  0.19  (9)  ; 
Mytilus  edulis  (high),  3.21  ±  0.61  (6)  ;  Mercenaria  mercenaria,  6.69  ±  0.25  (10)  ; 
Modiolus  modiolus,  5.08  ±  0.24  (7)  ;  Spisula  solidissima,  11.40  ±  0.67  (6)  ;  Myti- 
lus edulis  (low),  5.18  ±  0.45  (8)  ;  and  Placopectcn  imujclhmlcits,  11.29  ±  0.07  (5) 
X  10~5  cm/sec.  In  addition  the  diffusional  water  permeability  for  the  isolated 
mantle  tissue  of  a  representative  freshwater  bivalve,  Anodonta  sp.,  was  found  to  be 
6.17  ±  0.22  (7)  X  10-5  cm/sec. 

Measurements  of  the  thickness  of  the  mantles  used  in  some  of  these  perme- 
ability determinations  are  given  in  Table  I.  There  is  apparently  no  correlation 
between  the  thickness  of  the  mantle  tissue  and  its  measured  water  Pa.  For  ex- 
ample, there  is  no  real  significant  difference  between  the  thickness  of  mantle  tissue 
used  from  Mytilus  edulis,  Modiolus  dcuiissus  and  Modiolus  modiolus  in  these 
studies,  although  there  is  a  considerable  difference  in  their  water  permeabilities. 
This  apparent  independence  between  tissue  thickness  and  water  permeability  may 
be  accounted  for  if  the  tissue  itself  is  nonhomogeneous  as  far  as  water  movements 
are  concerned.  That  is,  permeability  is  a  positive  function  of  thickness  in  a  given 
system  only  if  the  resistance  to  diffusive  movement  is  constant  throughout  its  en- 
tire thickness.  Alternatively,  one  thin  rate  limiting  diffusion  barrier  may  be  in 
series  with  the  thicker,  more  permeable  tissue.  In  this  case,  the  measured  tissue  P(] 
would  be  apparently  independent  of  the  thickness  of  the  tissue  (Prusch  and  Benos, 
1976). 

An  interesting  question  arises  from  this  study  concerning  the  ability  of  intertidal 
bivalves  to  alter  their  tissue  water  permeability  in  response  to  osmotic  stress,  such 
as  has  been  previously  reported  in  some  crustaceans  (Herreid,  1969a).  In  Figure 
2,  the  mantle  water  P(I  is  presented  for  two  groups  of  Mytilus  edulis  from  the  same 
general  area,  one  collected  intertidally  and  the  other  found  sublittorally.  Those 
animals  found  higher  intertidally  had  a  significantly  lower  tissue  water  P(1  than 
those  found  in  the  sublittoral  zone.  This  would  suggest  either  that  the  two  groups  of 
mussels  represent  different  physiological  races  or  that  this  animal  is  capable  of 
altering  its  tissue  permeability  in  response  to  an  osmotic  or  desiccation  stress  situa- 


BIVALVE  WATER  PERMEABILITY 


297 


1  U 

8 

o 

\ 

6 

E 

0 

in 

'o 

4 

X 

0? 

2 

8 


CONTROL 


o 
o 


o 
o 


o 
o 


o 
o 


70% 
SEAWATER 


•r 


0 

4 

8 

12 

16 

20 

24 

28 

32 

36 

DAYS 

FIGURE  3.  Diffusional  water  permeability  as  a  function  of  time  in  Mytilus  cditlis;  solid 
points  represent  animals  maintained  in  full  strength  sea  water;  and  open  points;  animals 
maintained  in  70%  sea  water. 

tion.  In  experiments  with  a  different  population  of  Mytilus  edulis  in  the  Woods 
Hole  area,  it  was  found  that  these  animals  could  indeed  decrease  their  tissue 
permeability  in  response  to  an  altered  osmotic  environment  (70%  sea  water)  as  is 
shown  in  Figure  3.  Those  animals  maintained  in  full-strength  sea  water  maintained 
a  constant  mantle  water  Pd  for  the  duration  of  the  experiment,  35  days,  at  about 
6  X  10  D  cm/sec,  while  those  in  70%  sea  water  slowly  decreased  their  mantle  water 
P,i  during  this  time  period  to  about  2  X  10  5  cm/sec. 

DISCUSSION 

Diffusional  water  permeabilities  in  a  series  of  lamellibranchs,  as  measured  across 
the  isolated  mantle  tissue,  can  be  correlated  with  the  habitat  of  the  animal  (Fig.  2). 
The  animals  used  in  this  study  can  be  divided  into  several  groups:  high  intertidal 
or  estuarine,  low  intertidal,  and  sublittoral.  In  general,  the  diffusional  water  perme- 
ability of  the  mantle  decreases  with  increasing  osmotic  or  desiccation  stress  to  which 
the  animal  is  exposed.  That  is,  Modiolns  deiuissits  and  C.  virginica,  which  are 
found  high  in  the  salt  marsh,  are  much  less  permeable  to  water  than  are  P.  inagcl- 
lanicHS  and  S.  solidissiina,  which  are  located  in  relatively  deep  water.  Organisms 
which  are  exposed  intermittently  or  for  brief  periods  of  time,  such  as  Mya,  Merccn- 
aria  or  M.  modiolns,  have  intermediate  diffusional  water  permeabilities. 

In  this  particular  study,  two  different  groups  of  Mytilus  edulis  were  used. 
One  group  was  distributed  on  rocks  low  in  the  intertidal  zone,  approximately 
ELWX,  while  another  group  at  the  same  site  (Jamestown.  Rhode  Island)  was 
found  sublittorally.  As  can  be  seen  in  Figure  2,  water  Pd  in  the  isolated  mantle 
tissue  from  these  two  groups  of  animals  is  significantly  different,  with  animals  from 
the  higher-zoned  area  having  the  lower  water  permeability.  This  not  only  indi- 


298  K.   H.  PRUSCH  AND  C.  HALL 

catcs  that  difiusional  water  permeability  in  bivalve1  molluscs  is  influenced  by  vertical 
zonation  level,  but  that  some  bivalves  may  be  capable  of  altering  the  water  perme- 
ability of  their  tissues  in  response  to  increased  environmental  osmotic  or  desiccation 
stress  situations.  Changes  in  water  permeability  in  response  to  increased  osmotic 
stress  has  already  been  demonstrated  in  several  crustaceans  (Herreid,  1969a; 
Lockwood  ct  «/.,  1973)  and  in  Linntlits  (Hannan  and  Evans,  1973). 

The  possibility  of  short  term  changes  in  tissue  water  \\\  in  bivalves  was  investi- 
gated in  what  presumably  was  a  homogenous  population  of  Mytilus  cdulis  collected 
from  deep  water  off  of  Martha's  Vineyard.  These  animals  were  split  into  two 
groups,  one  maintained  in  sea  water  and  the  other  in  70%  sea  water.  Those 
animals  maintained  in  full-strength  sea  water  maintained  a  constant  water  P,i, 
while  those  animals  in  70%  sea  water  demonstrated  a  slow,  steady  decrease  in 
tissue  water  permeability  which  leveled  off  at  a  new  steady  state  value  (2  X  10~5 
cm/sec)  approximately  24-28  days  after  the  initiation  of  the  equilibration  period 
(Fig.  3).  This  indicates  that  tissue  water  permeability  can  be  altered  in  at  least 
Mytilits  cdulis.  Since  Mytilus  edulis  is  incapable  of  any  significant  degree  of  ionic 
or  osmotic  regulation  (Potts,  1954),  then  what  is  the  physiological  significance  of 
this  permeability  change?  It  may  simply  be  that  those  specimens  of  Mytilus,  which 
are  located  in  areas  subjected  to  alterations  in  the  osmotic  and  ionic  environment, 
decrease  the  rate  of  tissue  equilibration  in  response  to  these  environmental  altera- 
tions by  decreasing  their  tissue  permeability. 

Anodonta,  a  freshwater  lamellibranch,  has  a  diffusional  water  permeability 
which  is  also  intermediate  between  high  and  subtidal  species.  In  a  similar  study 
with  crustaceans,  Rudy  (1967)  found  that  Astacns,  a  freshwater  crayfish,  had  the 
lowest  water  permeability  found  in  a  series  of  decapod  crustaceans  ranging  from 
marine  to  freshwater  species.  Why  this  is  not  the  case  in  a  similar  series  of 
lamellibranchs  used  in  this  study  is  not  known,  but  may  be  related  to  the  mainte- 
nance of  an  extremely  low  osmolality  of  the  hemolymph  in  these  organisms  (Potts, 
1954).  Since  these  animals  maintain  themselves  only  slightly  hyperosmotic  to  their 
environment  in  comparison  with  other  freshwater  invertebrates,  their  osmotic  prob- 
lems are  correspondingly  decreased,  and  they  may  not  need  to  reduce  their  tissue 
water  permeability  further  in  order  to  maintain  their  osmotic  equilibrium. 

Without  regard  to  the  actual  mechanism  by  which  water  moves  across  the  iso- 
lated mantle,  but  assuming  the  mechanism  of  water  movement  across  the  mantle 
tissue  is  the  same  in  the  different  bivalves  used  in  this  study,  reduction  of  tissue 
water  permeability  could  be  brought  about  in  one  of  several  different  means.  These 
would  include  reduction  of  exposed,  permeable  surfaces,  increasing  tissue  thickness 
or  changes  in  the  chemical  composition  of  the  tissue,  among  other  possibilities.  An 
organism  could  reduce  the  total  surface  area  of  permeable  tissues,  thereby  reducing 
its  overall  permeability.  This  mechanism  has  been  utilized  by  some  crustaceans  in 
which  it  has  been  noted  that  there  is  a  reduction  in  gill  area  per  unit  weight  going 
from  subtidal  to  intertidal  species,  the  gill  being  the  most  permeable  structure  in 
crustaceans  (Gray,  1957).  This  noted  reduction  in  crustacean  gill  area  may  also 
be  dependent  in  part  upon  the  availability  of  oxygen.  That  is,  proceeding  from 
the  subtidal  to  the  estuarine  and  terrestrial  environments  there  is  an  increase  in 
the  availability  of  oxygen,  and  therefore  an  animal  could  carry  out  its  respiratory 


BIVALVE  WATER  PERMEABILITY  299 

functions  with  less  gill  surface  area.  Intertidal  bivalves  have  most  likely  not  re- 
sorted to  a  reduction  in  gill  surface  area  as  a  means  to  decrease  their  total  surface 
permeability,  even  though  the  gills  make  a  major  contribution  to  the  total  exposed 
surface  area  in  these  animals.  The  gill  in  these  animals  is  used  for  filter-feeding, 
as  well  as  maintaining  a  respiratory  function,  and  with  a  reduction  in  the  time 
available  for  this  type  of  feeding  higher  in  the  intertidal  area,  reduction  in  gill  sur- 
face area  would  most  likely  be  counter-productive. 

Alternatively,  an  organism  could  reduce  the  diffusional  water  permeability  of  a 
given  tissue  by  increasing  the  thickness  of  the  tissue.  In  a  series  of  different  bi- 
valve mantles,  there  was  no  correlation  between  measured  mantle  thickness  and  dif- 
fusional water  permeability  (Table  I).  In  addition,  there  was  no  change  in  lag 
time  (Fig.  1)  across  the  mantles  with  different  permeabilities,  which  is  what  could 
be  expected  if  decreases  in  P,i  were  brought  about  by  increased  tissue  thickness. 
Apparently  then,  molluscs  have  not  utilized  this  possibility  to  decrease  tissue 
permeability. 

Without  changes  in  the  physical  dimensions  of  a  given  tissue,  changes  in  dif- 
fusional water  permeability  could  be  brought  about  by  changes  in  the  chemical 
composition  of  the  tissue.  For  example,  the  water  permeability  of  artificial  bilayer 
membranes  is  influenced  by  lipid  composition  (Cass  and  Finkelstein,  1967;  Granzi- 
ani  and  Livne,  1972).  Exposure  of  the  blue  crab  Callinectcs  to  decreased  osmo- 
lality  results  in  an  increase  in  lipid  synthesis  in  the  gill  (Whitney,  1974).  Callinectes 
is  an  estuarine  organism  and  is  capable  of  withstanding  large  changes  in  external 
salinity,  accomplished  in  part  apparently  by  decreasing  its  water  permeability.  The 
spider  crab  Libinia  on  the  other  hand  is  a  sublittoral  animal  incapable  of  any  great 
degree  of  osmoregulation.  When  this  animal  was  exposed  to  lowered  salinities, 
there  was  no  change  in  gill  lipid  synthesis.  Differences  in  diffusional  water  perme- 
ability across  the  mantle  tissue  of  bivalves  from  different  habitats  may  then  reflect 
differences  in  the  lipid  composition  of  the  tissue.  That  is,  there  may  be  an  increase 
in  the  lipid/protein  ratio  in  the  mantle  tissue  with  increasing  exposure  to  osmotic 
stress  resulting  in  decreased  diffusional  water  permeability. 

Adaptations  to  water  problems  in  other  intertidal  molluscs  include  structural, 
behavioral,  and  physiological  processes.  The  structure  of  the  shell  in  the  European 
limpet  Patella  is  correlated  with  their  intertidal  distribution  (Davies,  1969).  Higher 
zoned  animals  have  higher  shells  with  a  smaller  circumference  than  lower  zoned 
animals.  This  effectively  reduces  the  surface  area  of  the  higher  zoned  animals. 
Davies  also  suggested  that  there  may  be  differences  in  the  water  permeability  of  the 
mantle  tissue  of  these  limpets.  The  false  limpet  Siphonaria  pectimita  has  no  ability 
to  osmoregulate  but  can  tolerate  salinities  between  20  and  409r-  Salinity  variations 
outside  of  the  tolerance  range  cause  the  animal  to  contract  the  foot  musculature 
creating  a  seal  between  the  shell  and  substrate,  effectively  shutting  out  the  external 
environment  (McAlister  and  Fisher,  1968).  Wolcott  (1973)  claims  that  the  most 
important  adaptation  of  Acmaca  digitalis  to  environmental  stress  situations  is  the 
secretion  of  a  mucous  sheet  between  the  shell  margin  and  substratum,  again  sealing 
off  the  external  environment,  a  situation  analogous  to  the  secretion  of  the  epiphragm 
in  terrestrial  snails  (Alachin,  1968). 

Although  biotic  factors,  such  as  competition,  behavior,  predation,  etc.,  probably 


3()0  K.  I).   PRUSCH  AND  C.  HALL 

play  a  major  role  in  the  vertical  distribution  of  intertidal  animals  (Wolcott,  1973), 
abiotic  factors  also  influence  the  intertidal  distribution  of  these  animals  (Newell, 
1970).  This  present  study  suggests  that  the  ability  of  certain  groups  of  animals 
to  adapt  physiologically  to  specific  environmental  stress  situations  may  also  influ- 
ence their  distribution. 


Supported  in  part  by  grant  NS-09090. 

SUMMARY 

1.  The  diff visional  water  permeability  of  the  isolated  mantle  tissue  from  a  series 
of  marine,  and  one  freshwater,  species  of  lamellibranch  molluscs  was  determined. 

2.  The  water  permeability  of  the  mantle  tissue  was  generally  correlated  with 
the  habitat  of  the  organism,  permeability  decreasing  with  increasing  height  above 
the  sublittoral  zone. 

3.  Evidence  is  presented  that  a  given  intertidal  lamellibranch  species,  Mytilus 
edulis,  is  capable  of  altering  its  tissue  water  permeability  when  presented  with 
changes  in  external  osmolality. 

4.  The  observed  differences  in  tissue  water  permeability  from  different  animals 
are  not  due  to  change  in  the  physical  dimensions  of  the  tissue,  but  may  be  the  result 
of  changes  in  the  chemical  composition  of  the  tissue. 

LITERATURE  CITED 

BROEKHUYSEN,  C.  J.,  1940.  A  preliminary  investigation  of  the  importance  of  desiccation,  tem- 
perature and  salinity  as  factors  controlling  the  vertical  distribution  of  certain  marine 
gastropods  in  False  Bay,  South  Africa.  Trans.  R.  Soc.  S.  Ajr.,  28 :  255-292. 

BROWN,  A.  C.,  1960.  Desiccation  as  a  factor  influencing  the  vertical  distribution  of  some  South 
African  Gastropoda  from  intertidal  rocks  shores.  Port.  Acta  Biol.  Scr.  A,  7:  11-23. 

CASS,  A.,  AND  A.  FINKELSTEIN,  1967.  Water  permeability  of  thin  lipid  membranes.  /.  Gen. 
Physio!..  50:  1765-1784. 

DAVIES,  P.'S.,  1969.  Physiological  ecology  of  Patella.  III.  Desiccation  effects.  /.  Mar.  Biol. 
Assoc.  U.K.,  49:  291-304. 

FOSTER,  B.  A.,  1971.  Desiccation  as  a  factor  in  the  intertidal  zonation  of  barnacles.  Mar.  Biol., 
8 :  12-29. 

GRAY,  I.  E.,  1957.    A  comparative  study  of  the  gill  area  of  crabs.  Biol.  Bull.,  112  :  34-42. 

GRAZIANI,  Y.,  AND  A.  LIVNE,  1972.  Water  permeability  of  bilayer  lipid  membranes:  sterol- 
lipid  interaction.  J.  Mcwbr.  Biol.,  7  :  275-284. 

GROSS,  W.  J.,  1957.  An  analysis  of  response  to  osmotic  stress  in  selected  decapod  Crustacea. 
Biol.  Bull.,  112:  43-62. 

HANNAN,  J.  F.,  AND  D.  H.  EVANS,  1973.  Water  permeability  in  some  euryhaline  decapods  and 
Limulus  Polyphemus.  Coinp.  Biochcm.  Physiol.,  44A  :  1199-1213. 

HERREID,  C.  F.,  1969a.  Integument  permeability  of  crabs  and  adaptation  to  land.  Cuinp.  Bio- 
chcm. Physiol.,  29 :  423-429. 

HERREID,  C.  F.,  1969b.  Wrater  loss  of  crabs  from  different  habitats.  Comp.  Biochcm.  Physiol., 
28:  829-839. 

ISTIN,  K.,  AND  L.  KIRSCHNER,  1968.  On  the  origin  of  the  bioelectric  potential  generated  by  the 
fresh  water  clam  mantle.  /.  Gen.  Physiol.,  51  :  478-496. 

KROGH,  A.,  1939.  Osmotic  rcunlatitm  in  aquatic  anunalx.  Cambridge  University  Press,  London, 
242  pp. 


BIVALVE  WATER  PERMEABILITY  301 

LOCKWOOD,  A.  P.  M.,  1962.    The  osmoregulation  of  Crustacea.    Biul.  Rev.,  37:  257-305. 
LOCKWOOD,  A.  P.  M.,  C.  B.  E.  INMAN,  AND  T.  H.  COURTEXAY,   1973.     The  influence  of  en- 
vironmental salinity  on  the  water  fluxes  of  the  amphipod  crustacean  Gammarus  ducbcni. 

J.  Ex  p.  Biol,  58  :  137-148. 
MACHIN,   J.,   1968.     The   permeability  of   the   epiphragm   of  terrestrial   snails   to   water   vapor 

Biol  Bull.,  134:  87-95. 
McALisTER,  R.  O.,  AND  F.  M.  FISHER,  1968.     Response  of  the  false  limpet,  Siphonaria  pectinata 

Linnaeus   (Gastropoda,  Pulmonata)   to  osmotic  stress.     Biol.  Bull,  134:  96-117. 
NEFF,  J.,   1972.     Ultrastructure  of  the  outer  epithelium  of  the  mantle  in  the  clam  Mercenaria 

mercenaria  in  relation  to  calcification  of  the  shell.     Tissue  and  Cell,  4:  591—600. 
NEWELL,  R.  C.,  1970.     Biology  of  iiitcrtidal  animals.    American  Elsevier  Publishing  Co.,  New 

York,  555  pp. 

PILGRIM,  R.  L.  C.,  1953.     Osmotic  relations  in  molluscan  contractile  tissues.     I.   Isolated  ven- 
tricle-strip   preparations    from    Lamellibranchs    (Mytilns    cdulis   L.,    Ostrea    edulis   L., 

Anodonta  cygnea  L.).    /.  Exp.  Biol,  30:  297-316. 
POTTS,  W.  T.  W.,  1954.     The  inorganic  composition  of  the  blood  of  Mytilus  cdulis  and  Ano- 

donta  cygnea.   J.  Exp.  Biol,  31 :  376-385. 
PRUSCH,  R.  D.,  AND  D.  J.  BENOS,  1976.     Cuticular  control  of  diffusional  water  permeability. 

/.  Insect  Physio!.,  22  :  629-632. 
RUDY,  P.  P.,  1967.    Water  permeability  in  selected  decapod  Crustacea.   Comp.  Biochem.  Physiol, 

22  :  581-589. 
SMITH,  R.  L,  1970.    The  apparent  water  permeability  of  Carcimts  macnas  (Crustacea,  Brachy- 

ura,  Portunidae)  as  a  function  of  salinity.  Biol  Bull,  139:  351-362. 
VERNBERG,  F.  J.,  C.  SCHIEPER,  AND  D.  E.  SCHNEIDER,  1963.     The  influence  of  temperature  and 

salinity  on  ciliary  activity  of  excised  gill  tissue  of  molluscs  from  North  Carolina.    Comp. 

Biochem.  Physiol,  8  :  271-285. 
WHITNEY,   J.,    1974.     The  effect   of  external    salinity   upon   lipid   synthesis    in   the   blue   crab 

Callincctes  sapidus  Rathbun  and  in  the  spider  crab  Libinia  cnwrginata  Leech.     Comp. 

Biochem.  Physio!.,  49A  :  433-440. 
WOLCOTT,   T.,    1973.      Physiological   ecology   and   intertidal   zonation   in   limpets    (Acmaca)  :   a 

critical  look  at  limiting  factors.  Biol  Bull.,  145:  389-422. 


Reference:   li'wl.  Hull..  154  :   M2-321.    (April,  1978) 


THE  LIGHT-DARK  CYCLE  AND  A  NONLINEAR  ANALYSIS  OF 

LUNAR  PERTURBATIONS  AND  BAROMETRIC  PRESSURE 

ASSOCIATED  WITH  THE  ANNUAL  LOCOMOTOR 

ACTIVITY  OF  THE  FROG,  KANA  P1PIENS 

DOUGLAS  R.  ROBERTSON 

Department  of  Anatomy,  State  University  of  New  York,  Upstate  Medical  Center, 

Syracuse,  Neiv  York  13210 

During  the  course  of  study  on  intestinal  calcium  transport  in  the  frog  Kana 
pipicns,  it  became  apparent  that  maximal  transport  activity  in  April  to  June  oc- 
curred during  the  nocturnal  hours  (Robertson,  1976).  This  period  of  increased 
physiological  activity  of  the  digestive  system  may  be  related  to  feeding  behavior,  but 
at  present  few  studies  have  been  made  to  determine  the  activity  patterns  of  the 
leopard  frog.  Observations  of  frogs  as  a  group  in  temperate  zones  show  that 
activity  is  present  during  the  day  as  well  as  at  night,  depending  upon  the  species 
(Wright  and  Wright,  1949).  At  lower  latitudes,  as  in  Panama,  Bufo  inarinits  is 
active  primarily  during  the  day  (Park,  Barden  and  Williams,  1940)  ;  however, 
Jaeger,  Haitman  and  Jaeger  (1976)  noted  in  the  Panamanian  frog,  Colostethus, 
a  bimodal  activity  pattern  with  maximums  at  0830  and  1630  hr.  This  activity 
pattern  was  coincident  with  the  crepuscular  activity  of  the  primary  food  source, 
dipterans  and  coleopterans. 

With  respect  to  the  Ranidae  group,  Rana  temporaria  displays  higher  locomotor 
activity  at  night  (Chugunov  and  Kuznetsov,  1972),  while  Rana  escnlcnta  is  active 
throughout  the  day  (Kuznetsov,  Chugunov  and  Brodskii,  1972).  Dole  (1972) 
noted  that  newly  metamorphosed  specimens  of  Rana  pipicns  were  more  active 
after  nocturnal  rains.  The  food  gathering  activity  patterns  of  adult  Rana  pipicns 
may  be  inferred  from  the  stomach  contents  which  contain  those  insects  most  com- 
monly available  at  various  times  of  the  feeding  season  (April  to  October)  (Whit- 
aker,  1961;  Linzey,  1967).  Diet  appears  to  be  a  reflection  of  the  availability  of 
various  insects  rather  than  preference  and  includes  species  of  Coleoptera  (beetles, 
weevils),  Hymenoptera  (ants),  and  Homoptera  (aphids,  leafhoppers).  Of  these, 
beetles  (Oldryd,  1960)  and  ants  (Skaife,  1961)  display  increased  activity  at  night. 

When  food  is  made  available  at  a  preset  period  of  time,  the  bullfrog,  Rana 
catesbeiana,  exhibits  increased  activity  presumably  in  anticipation  of  food  avail- 
ability, suggestive  of  a  "conditioned"  response  (van  Bergeijk,  1967). 

In  spontaneous  or  nonconditioned  activity,  amphibians  appear  to  be  sensitive 
not  only  to  the  general  environment,  such  as  season  and  rainfall  (Gibbons  and 
Bennett,  1974),  but  to  solar  and  lunar  cues  for  orientation  and  migration  (Fergu- 
son, Landreth  and  Turinipseed,  1965;  Ferguson  and  Landreth,  1966;  Landreth 
and  Ferguson,  1966,  1967;  FitzGerald  and  Bider,  1974).  When  maintained  in  a 
closed  environment  away  from  visual  cues,  the  activity  of  salamanders  (Trit tints ) 
appears  to  be  modified  by  a  lunar  influence  (Ralph,  1957),  which  may  also  in- 
fluence physiological  activity,  such  as  oxygen  consumption  (Brown,  Webb,  Bennett 

302 


LOCOMOTOR  ACTIVITY  OF  FROG  303 

and  Sanders.  1955).  In  an  attempt  to  provide  an  overview  of  the  diurnal  activity 
of  physiological  processes  of  the  digestive  system,  a  year-long  study  was  conducted 
on  the  spontaneous  activity  patterns  of  adult  male  Rana  pipiens  exposed  to  the 
change  in  outdoor  ambient  lighting  conditions. 

MATERIAL  AND  METHODS 

Adult  male  specimens  of  Rana  pipiens  (Northern  variety)  of  40-60  g  body 
weight  \vere  obtained  commercially  (Bay  Biological,  Canada)  throughout  the  period 
of  study  from  March,  1976,  to  February,  1977,  with  studies  conducted  in  Syracuse, 
Xew  York  (Lat.  43°  03'  N).  Frogs  were  unfed  and  maintained  in  a  continuous 
change  of  fresh  water  (20-24°  C)  and  exposed  to  outdoor  ambient  diurnal  lighting 
conditions  for  one  week  prior  to  use.  After  this  period,  groups  of  8-10  frogs  were 
placed  in  the  detection  apparatus  described  below  in  a  relatively  quiet  room.  The 
frogs  were  exposed  to  southern  ambient  lighting  (300-400  lux  max).  The  sur- 
rounding yearly  temperature  was  maintained  at  19.3  ±3.1°  C  and  only  significantly 
elevated  above  the  yearly  mean  in  September.  Xewly  acclimated  frogs  were  intro- 
duced into  the  apparatus  at  irregular  intervals  after  10-20  days.  Since  Brown, 
Webb  and  Macey  (1957)  had  noted  that  barometric  pressure  may  affect  biological 
activity  in  amphibians,  local  barometric  pressure  at  1200  hr  was  recorded  on  a 
recording  aneroid  barometer  throughout  the  course  of  study. 

Detection  apparatus 

The  spontaneous  locomotor  activity  of  groups  of  frogs  was  monitored  by  the 
detection  of  vertical  wrater  movement  in  an  isolated  translucent  plastic  chamber  (33 
X  23  X  10  cm).  The  apparatus  consisted  of  a  plastic  float  connected  to  a  pivoted 
transverse  rod  through  which  a  vertical  water  displacement  of  2.0  ±  0.25  mm  was 
multiplied  by  a  factor  of  five.  A  contact  switch  at  the  end  of  the  rod  completed  a 
circuit  with  an  event  marked  on  a  strip  recorder.  The  switch  interval  was  ad- 
justed daily  to  maintain  a  constant  sensitivity. 

Analysis  of  data 

Locomotor  activity  (LA)  was  recorded  as  the  number  of  events/hour  and  the 
number  of  events/day.  Hourly  variations  were  reduced  by  averaging  the  data  for 
two  hour  periods.  Further,  the  relative  activity  (per  cent  activity/two  hour  period) 
was  calculated  from  the  total  number  of  events/day.  Since  lunar  perturbations  ap- 
peared to  influence  amphibian  behavior  (Ralph,  1957),  identification  of  coherent 
patterns  of  locomotor  activity  was  facilitated  by  the  construction  of  isograms  of  the 
relative  levels  of  activity  within  a  lunar  month.  The  mean  relative  activity  pattern 
for  a  "typical  lunar  month"  was  calculated  by  deriving  the  mean  relative  activity 
(per  cent/two  hours)  of  corresponding  lunar  days  over  a  three  to  four  lunar 
month  period.  Lines  were  connected  between  equal  levels  of  activity  (5%  inter- 
vals) to  provide  a  "contour  map"  which  would  emphasize  basic  activity  patterns. 

The  presence  of  a  lunar  influence  during  the  diurnal   and   nocturnal  periods 
was  determined  by  a  procedure  described  by  Brown  ct  <//.   (1955).     High  transit 


304 


DOUGLAS  R.  ROBERTSON 


(JIT;  time  of  moon  at  zenith)  advances  50  minutes  each  day,  and  completes  a  cycle 
in  about  2^.53  days  (synodical  lunar  month).  The  absolute  levels  of  locomotor 
activity  at  HT  for  each  day  were  then  tabulated  into  a  single  column  (Column  3) 
with  the  data  of  the  subsequent  time  period  shifted  accordingly  to  the  right.  This 
procedure  enhances  any  lunar  effect  associated  with  the  time  of  HT  that  is  super- 
imposed upon  the  general  level  of  activity.  The  times  of  transit  were  obtained  from 
the  American  Ephemcris  and  Nautical  Almanac  (1976,  1977,  pp.  52-67),  and  all 
time  periods  designated  in  Eastern  Standard  Time.  Day  of  lunar  month  is  desig- 
nated as  (NM  +  day). 

Period  analysis 

A  computer  program  was  designed  to  identify  and  analyze  the  times  of  maximal 
activity  with  respect  to  a  specific  reference  date.  Since  most  activity  was  during 
the  nocturnal  period,  the  time  of  maximal  locomotor  activity  (LAmax)  was  recorded 
as  the  time  (hours)  before  (  — )  or  after  (  +  )  0000  hr.  For  example,  a  LAmax 
period  of  activity  at  2000  hr  was  recorded  as  —4  (hr),  while  an  activity  peak  at 
0400  hr  was  recorded  as  +4  (hr).  Initial  observations  suggested  that  the  time  of 
LAmax  occurred  at  a  different  time  each  day  but  was  repeated  several  days  later  at  a 
similar  time  period.  This  could  best  be  visualized  as  an  oscillatory  cycle  with  the 
function:  LAmax  =  M  +  cos  (o>t  —  0).  where  the  cyclic  parameters  are  M  (mean  of 
all  LAmax  time  values),  A  (amplitude  of  the  oscillation),  and  <£  (arcophase  or  phase 
angle,  which  is  the  day  in  the  cycle  in  which  LAniax  occurs  at  the  maximal  positive 
hour  after  0000  hr),  after  a  specific  reference  date  expressed  in  degrees.  The  refer- 
ence date  was  taken  as  day  1  (NM  +  1)  of  a  lunar  month  on  4/1/76.  The  angu- 
lar frequency  o>  —  360 °/t,  where  t  --  number  of  days  in  the  cycle.  For  cyclic 


TABLE  I 

Monthly  locomotor  activity  levels  of  adult  male  frogs,  Rana  pipiens. 


Month 

N 
(Days) 

Events/day  ±  s.e.m. 

Temp,  range  °  C 
(min-max) 

Barometric  pressure 
(mm  Hg)* 

January 

17 

539  ±  78 

15.0-18.0 

756.4  ±  7.2 

February 

28 

420  ±  60 

15.5-18.0 

759.5  ±  7.3 

March 

11 

297  ±  14 

15.0-17.0 

761.2  ±6.0 

April 

28 

373  ±  26 

16.0-18.5 

762.3  ±  3.3 

May 

30 

387  ±  52 

18.0-20.0 

760.0  ±  5.8 

June 

21 

405  ±  50 

20.0-23.0 

762.8  ±4.2 

July 

10 

286  ±  30 

22.0-23.0 

760.2  ±  4.4 

August 

29 

281  ±  34 

22.0-23.0 

763.3  ±  4.0 

September 

29 

211  ±  36 

23.0-26.0 

762.5  ±  5.2 

October 

31 

256  ±  29 

20.0-22.0 

762.0  ±  6.8 

November 

30 

146  ±  31 

17.0-20.0 

760.5  ±  5.7 

December 

14 

159  ±  45 

15.0-18.0 

759.7  ±  8.7 

Total  annual 

mean 

278 

313  ±  15 

19.4  ±  3.1 

761.0  ±6.1 

Mean  pressure  at  1200  hr. 


LOCO  MOTOR  ACTIVITY  OF  FROG 


305 


analysis  of  LA,,,ax.  the  time  of  maximal  activity  was  paired  with  the  number  of  the 
day  after  the  reference  date.  The  data  was  then  tested  with  the  cosine  function 

J 

above  with  hypothetical  wave  functions  of  various  periods  where  t  --  5  to  90  days 
for  the  monthly  analysis  and  cycles  up  to  360  days  for  all  data  collected  for  the 
entire  year.  The  procedure  was  conveniently  done  through  a  linear  transformation 
of  the  cosine  time  parameter  in  which  the  data  was  analyzed  as  a  simple  linear 
regression.  The  "best  fit  of  the  data"  was  expressed  as  the  maximum  positive 
value  of  the  correlation  coefficient  (r)  at  a  specific  <£>.  This  value  at  <f>  was  obtained 
by  calculating  the  values  of  r  at  10°  increments  through  the  entire  wave  function  of 
360°  for  each  hypothetical  cycle.  The  resulting  maximum  values  of  r  for  the  range 
of  cycles  represents  a  spectrum  in  which  the  t  statistic  was  calculated  and  P  <  0.01 
was  considered  a  significant  cycle.  Possible  correlations  of  barometric  pressure  and 
the  daily  absolute  levels  of  locomotor  activity  on  the  cyclic  spectrum  were  con- 
veniently analyzed  by  processing  the  data  only  for  those  days  when  the  values  were 
above  or  below  the  annual  mean.  All  data  expressed  in  the  text  are  mean  ±  s.d. 
unless  otherwise  stated,  and  P  <  0.01.  determined  from  a  simple  /-statistic  between 
values,  was  considered  significant. 


1200 


20      22       24      26      28 


LUNAR    DAY 


Locomotor  Activity  -  3  Mo. 
APRIL  -  JUNE 


FIGURE  1.  Isogram  of  relative  locomotor  activity  (%  activity/two  hours)  during  April  to 
June  based  upon  the  mean  activity  levels  of  corresponding  lunar  days  over  a  three  lunar  month 
period.  A  coherent  sinusoidal  pattern  is  apparent  which  oscillates  within  the  lunar  month  to 
reflect  high  activity  at  0000  hr  near  New  Moon  which  shifts  to  1800  hr  at  a  time  of  Full  Moon. 
Additional  activity  at  0600  hr  at  Full  Moon  indicates  a  bimodal  diel  pattern.  Diagonal  lines 
represent  times  of  high  and  low  lunar  transit. 


306 


DOUGLAS    R.    ROBERTSON' 


RESULTS 


uclirily  fmllcrns 


Since  8-10  frogs  were  contiinioiisly  monitored,  the  locomotor  activity  (LA) 
in  this  study  is  group  activity  in  contrast  to  activity  of  individual  frogs.  Such 
activity  throughout  the  year  (total  of  278  days)  revealed  changes  in  the  relative 
levels  of  hourly  activity  from  month  to  month  which  did  not  appear  to  be  related 
to  the  absolute  monthly  activity  level  or  the  ambient  temperature.  The  yearly  mean 
temperature  was  maintained  at  19.4  ±3.1°  C  and  only  significantly  elevated  above 
the  mean  in  September  (Table  I).  The  mean  daily  activity  levels  in  the  calendar 
months  of  November,  December  and  January  showed  significant  deviations  from 
the  mean  daily  activity  levels  of  313  ±  115  events/day  for  the  entire  year.  No 
significant  correlation  existed  between  the  mean  daily  activity  for  each  month  and 
the  corresponding  mean  monthly  ambient  temperature  (r  =  -0.395)  or  the  pre- 
vailing mean  monthly  barometric  pressure  (r  =  -0.421). 

Chi  square  analysis  of  the  relative  hourly  activity  for  each  month  showed  no 
significant  variation  from  "random"  activity  (S.3c/e  activity/two  hours)  in  the 
months  from  October  through  March,  while  a  significant  nonrandom  light-dark 
response  pattern  of  activity  was  apparent  from  April  through  September.  Further 
analysis  of  this  nonrandom  period  based  upon  monthly  relative  activity  isograms 


200 


0600 


0000 


o 


I  800 


200 


10       12        14       16       18       20      22       24       26      28 


LUNAR    DAY 


JULY- OCTOBER 


FIGURE  2.    Isogram  of  relative  locomotor  activity  (%  activity/two  hours)   for  a  three  lunar 
month  period  from  July  to  October.     Elevated  activity   (>10%)    is  primarily  nocturnal  while 
>  15%  activity  is  observed  between  2100-0000  hr  and  0400-0600  hr  which  reflects  a  bimodal 
diel  activity  pattern.     Diagonal  lines  indicate  time  of  high  and  low  lunar  transit. 


LOCOMOTOR  ACTIVITY  OF  FROG 


307 


1200 


0600 


cr 

Z> 

o 

I 


0000 


1800 


1200 


0 


12 


16        18 


14 

O 

LUNAR    DAY 


20       22 

(L 


24       26       28 


OCTOBER- MARCH 


FIGURE  3.  Isogram  of  relative  locomotor  activity  (%  activity/two  hours)  from  October 
to  March  based  on  data  of  four  lunar  months.  Activity  is  randomly  distributed  throughout  the 
24  hr  period  each  day  with  no  significant  diurnal-nocturnal  pattern  evident.  Diagonal  lines 
indicate  times  of  high  and  low  lunar  transit. 

revealed  a  change  in  pattern  during  July.  Thus,  three  major  periods  were  identi- 
fied, the  first  nonrandom  period  from  April  through  June  (79  days),  a  second  non- 
random  period  from  July  to  the  first  New  Moon  in  Octoher  (90  days),  and  the 
third  "random"  period  from  New  Moon  in  October  through  March  of  the  following 
year  (109  days). 

The  composite  "lunar  month"  isogram  constructed  for  the  April  to  June  period 
(three  lunar  months)  revealed  a  coherent  sinusoidal  pattern  with  periods  of  in- 
creased LA  (>  10%/two  hours)  near  0000  hr  at  New  Moon  (Fig.  1).  This  pat- 
tern shifted  toward  early  evening  hours  during  the  progression  of  the  lunar  month 
to  exhibit  maximal  LA  at  1800  hr  at  Full  Moon  (NM  +  15).  In  addition,  a 
second  period  of  increased  activity  was  observed  at  0600  hr  which  indicated  a  bi- 
modal  diurnal  pattern.  During  the  remainder  of  the  month,  the  primary  activity 
pattern  returned  toward  0000  hr  and  continued  the  shift  to  0300  hr  at  the  end  of  the 
lunar  month.  Thus,  a  single  mode  of  activity  appeared  from  NM  +  16  to  NM  +  5 
of  the  following  lunar  month  and  a  bimodal  activity  pattern  from  NM  +  6  to  NM 
+  15. 

The  isogram  based  upon  a  composite  of  three  lunar  months  of  the  second  period 
(July  through  October)  revealed  a  prominent  bimodal  diel  pattern  which  was 
maximal  at  0400-0600  hr  with  a  minor  maximum  at  about  2200  hr  (Fig.  2).  The 
major  early  morning  maximum  coincides  with  the  time  of  sunrise  (0430-0600  hr) 


308 


DOUGLAS  K.  KOUKKTSHX 


o 

h- 

< 
cc 

Q 


0.60 


1200 


1800  0000 

TIME    (  hours) 


0600 


1200 


FIGURE  4.  Isogratn  of  relative  locomotor  activity  (%/two  hours)  in  adult  male  Rana 
f>if>icns  as  a  function  of  the  prevailing  L :  D  ratio  over  the  course  of  an  entire  year.  During 
winter  with  an  L:  D  ratio  <  1.0,  activity  is  not  defined  into  a  significant  diurnal-nocturnal  pat- 
tern although  a  nonsignificant  increase  is  observed  at  sunrise.  An  L:D  ratio  between  1.0-1.45 
is  coincident  with  a  significant  increase  in  nocturnal  activity  between  2100-0000  hr  and  during 
the  hours  at  sunrise.  Above  an  L :  D  ratio  of  1.5,  frogs  are  most  active  at  sunrise  with  little 
activity  (<  6%)  during  daylight  hours. 


during  this  time  of  year.  The  sinusoidal  pattern  of  the  previous  period  did  not 
appear  to  be  dominant. 

The  third  period  (October  through  March),  classified  as  random  activity,  was 
substantiated  by  the  composite  isogram  based  upon  four  lunar  months  (Fig.  3). 
No  coherent  daily  pattern  was  apparent  and  the  overall  pattern  was  characterized 
by  higher  levels  (>  10%/two  hours)  of  activity  present  during  the  daylight  hours, 
and  low  activity  levels  (<  5%/two  hours)  at  night. 

The  change  in  pattern  appeared  to  be  related  to  the  seasonal  alterations  in  the 
L :  D  ratio  as  depicted  in  Figure  4.  An  isogram  based  upon  the  relative  mean 
LA/two  hours  for  each  calendar  month  as  a  function  of  the  prevailing  L :  D  ratio 
revealed  that  random  LA  from  mid-October  to  March  occurred  when  the  L :  D 
ratio  was  <  1.0.  As  the  L:  D  ratio  increased  to  1.0,  increased  LA  was  more  ap- 
parent at  time  of  sunrise  and  became  maximal  between  2100  and  0000  hr  up  to 
the  time  when  the  L:D  ratio  was  <  1.45.  Above  this  ratio  the  LA  was  pre- 
dominately in  the  hours  near  sunrise  with  <  6%  of  the  activity  occurring  during 
the  mid-daylight  hours. 

Lunar  influence 

The  relationship  of  lunar  position  on  hourly  LA  was  analyzed  by  columnating 
the  absolute  level  of  activity  for  each  two  hour  segment  during  the  time  of  High 


LOCOMOTOR  ACTIVITY  OF  FROG 


309 


P<0.05 

i        i        I        i I I 1 I 


JULY-  OCTOBER 


h- 
O 


cr 
o 
I- 
o 
^ 
o 
o 
o 


,-DIURNAL 
NOCTURNAL 


5 
10 


W    '2"? 


1 


DIURNAL 
NOCTURNAL 


I  I          I      I L 


_L 


DIURNAL 


rm 


IAL 


OCTOBER-MARCH 


23456789    10 

COLUMN    NUMBER 


12 


FIGURE  5.  Relationship  of  high  lunar  transit  (HT  in  column  3)  on  absolute  levels  of 
activity  (number  of  events/two  hours)  during  daylight  (open  circles  with  dashed  line)  and 
nocturnal  (solid  circles  with  solid  line)  hours  during  three  major  periods  of  the  year.  In  April 
to  June,  HT  during  nocturnal  period  had  no  significant  influence  on  activity  level  from  noc- 
turnal mean  (solid  line),  while  six  hours  after  HT  during  diurnal  period,  activity  was  sig- 
nificantly decreased  when  compared  to  daily  mean  (dashed  line).  In  July  to  October  mean 
diurnal  (dashed  line)  and  nocturnal  (solid  line)  activities  were  significantly  lower  than  activity 
in  preceding  and  subsequent  monthly  periods,  and  high  transit  was  not  associated  with  any 
significant  change  in  mean  hourly  activity.  In  October  to  March  mean  nocturnal  activity 
(solid  line)  was  significantly  below  mean  diurnal  activity  (dashed  line)  and  activity  four  to 
six  hours  prior  to  HT  during  daylight  was  significantly  above  diurnal  mean.  Each  point  is 
mean  ±  s.e.m. 


310  DOUGLAS  R.  ROBERTSON 

Transit  (I  IT)  (t'olunm  3  )  when  transit  occurred  cither  during  the  diurnal  or 
nocturnal  ])eriod  ( .Fig.  5).  During  April,  for  example,  I  IT  advanced  through  the 
nocturnal  period  in  14  days,  and  through  daylight  in  U>  days.  For  the  entire  April 
to  June  segment  the  time  of  HT  was  not  coincident  with  any  significant  change  in 
LA  at  any  time  during  the  dark  from  the  mean  nocturnal  level  of  20.4  ±  3.3  events/ 
two  hours.  When  transit  occurred  during  daylight  hours  there  was  a  significant 
decrease  for  six  hours  after  HT  below  the  diurnal  mean  of  17.3  ±  3.7  events/two 
hours.  Further,  the  mean  daylight  level  of  activity  was  significantly  below  the 
nocturnal  activity  level. 

High  transit  in  July  to  October  was  not  coincident  with  any  hourly  differential 
effect  on  diurnal  or  nocturnal  activity  levels,  nor  were  the  mean  activity  levels 
different  from  one  another  [(11. 3  ±1.9  (nocturnal)  vs.  10.9  ±  1.7  events/two 
hours  (diurnal)].  During  this  period,  the  total  number  of  events/day  (254  ±  17 
s.e.m.)  was  significantly  below  the  yearly  mean  of  313  ±  15  s.e.m.  events/day. 

The  association  of  HT  with  decreased  activity  observed  in  April-June  was 
reversed  during  the  period  from  October  to  March.  When  HT  occurred  during 
daylight  there  was  a  significant  increase  in  LA  six  hours  prior  to  transit  above  the 
mean  level  of  22.9  ±  3.9  events/two  hours  which  was  also  elevated  above  the  noc- 
turnal mean  of  20.2  ±  2.2  events/two  hours. 

Period  analysis 

The  characteristics  of  the  FORTRAN  Program  employed  in  this  study  are  such 
that  insignificant  values  of  r  will  be  generated  if  data  cannot  be  fitted  to  any  sine 
wave  by  least  squares  or  if  LAmax  is  a  linear  function  (e.g.,  LAmax  occurs  at  the 
same  time  each  day).  When  LAmax  was  analyzed  across  a  spectrum  of  hypothetical 
wave  lengths  at  five  day  intervals,  the  presence  of  significant  (P  <  0.01)  oscillatory 
patterns  appeared  as  peaks  of  increased  values  of  r.  This  procedure  performed  on 
each  of  the  three  monthly  segments  and  for  the  entire  year  revealed  the  presence  of 
cycles  indicative  of  coherent  and  organized  behavior  patterns.  Further  processing 
of  data  based  upon  the  relative  levels  of  barometric  pressure  and  locomotor  activity 
(events/day)  above  or  below  the  annual  mean  (Table  I)  altered  the  values  of  r 
(and  degree  of  significance)  of  specific  cycles.  Under  these  circumstances  the  phase 
angle  (<£)  was  identical  to  each  base  cycle  for  the  recalculation  of  r. 

In  April  to  June  (Fig.  6)  the  base  spectrum  for  all  data  indicated  the  presence 
of  a  28  day  and  50  day  cycle.  The  28  day  cycle  where  4>  —  0,  began  each  cycle  on 
the  first  day  of  New  Moon  during  this  period,  and  became  a  more  significant  cycle 
when  analyzed  for  those  days  of  low  barometric  pressure  (LBP)  ;  whereas  the  50 
day  cycle  which  began  14  days  into  the  lunar  month  was  more  prominent  under 
conditions  of  high  barometric  pressure  (HBP).  On  days  of  low  locomotor  activity 
(LLA)  a  slight  shift  in  the  cycle  occurred  with  the  increase  in  the  correlation 
coefficient  to  r  —  0.641  with  other  peaks  at  15  and  60  days. 

In  July  to  October,  two  additional  shorter  cycles  were  noted  in  addition  to  those 
seen  in  April  to  June  (Fig.  7).  The  5  day  cycle  was  more  significant  under 
conditions  of  LBP,  while  the  15  day  cycle  was  more  prominent  when  tested  on 
days  of  HBP.  As  noted  in  the  April  to  June  spectrum,  the  30  and  45  day  cycles 


LOCOMOTOR  ACTIVITY  OF   FROG 


311 


.6r— 


/'N 


Base 
HBP 
IBP 


(0°) 


I         (130°) 


30 


90 


HLA 
LLA 


60 


"A  (Days) 


April-June 


FIGURE  6.  Upper  graph  shows  a  period  analysis  of  times  of  LAma*  during  April  to  June  de- 
picted by  the  solid  line  indicating  two  significant  cycles  at  28  and  50  days.  The  resulting  spec- 
trum after  analysis  of  data  on  days  of  low  barometric  pressure  (LBP)  at  identical  tf>  shows 
enhancement  of  the  28  day  cycle,  while  on  days  of  high  barometric  pressure  (HBP)  the  50 
day  cycle  was  the  most  significant.  Lower  graph  shows  a  similar  analysis  on  days  in  which 
activity  was  lower  than  the  series  mean  (LLA)  a  30  day  cycle  was  most  significant,  whereas 
high  locomotor  activity  (HLA)  was  associated  with  the  50  day  cycle. 


were  more  prominent  under  conditions  of  HBP.  Examination  of  HLA  revealed  a 
significant  enhancement  of  the  15,  25,  and  50  day  cycles,  while  the  5  day  cycle  was 
dominant  on  days  of  LLA. 

For  the  period  from  October  to  March  (Fig.  8)  the  base  spectrum  displayed 
cycles  at  15,  30,  and  50  days  in  which  only  the  50  day  cycle  showed  a  significant 
enhancement  under  conditions  of  lower  than  average  barometric  pressure.  Higher 
than  average  locomotor  activity  was  coincident  with  the  15  and  65  clay  cycles 
while  LLA  was  associated  only  with  the  30  day  cycle. 

When  all  data  for  the  entire  year  was  examined  in  the  same  manner  (Fig.  9), 
only  the  significant  cycles  whose  phase  angle  was  relatively  constant  throughout  the 
year  could  be  detected.  The  significant  base  cycles  present  were  those  at  30,  55, 
105,  and  160  days.  On  days  of  HBP  the  105  and  K>()  day  cycles  were  most 


312 


DOUGLAS  R.  ROBERTSON 


6  i — 


.4 


.2 


30 


60 


90 


.6 


HLA 

LLA 


.4 


.2 


,'        I         ' 


'         I 


/• 

I: 


i    ; 


I   ' 


30 


60 


A  (Days) 


90 


July  -  October 


LOCOMOTOR  ACTIVITY  OF  FROG 


313 


TABLE  II 

Parameters  for  significant  monthly  and  annual  locomotor  activity  cycles. 


Time  segment 

Period 
(days) 

u 

<t> 

(Degrees) 

0 

(Day) 

M 

A 

Related  variables 
(see  text) 

April-June 

15 

24 

210 

9 

-0.73 

-1.98 

LLA 

20 

18 

270 

15 

-1.00 

-2.10 

HBP,  HLA 

30 

12 

30 

3 

-0.30 

3.35 

LBP,  LLA 

50 

7.3 

100 

14 

-0.95 

2.78 

HBP,  HLA 

July-October 

5 

72 

310 

4 

1.64 

-3.30 

LBP,  LLA 

15 

24 

170 

7 

1.08 

2.61 

HBP,  HLA 

27 

13.3 

120 

9 

1.15 

2.05 

LBP,  HLA 

43 

8.4 

220 

26 

3.38 

-2.09 

LLA 

50 

7.3 

310 

43 

2.41 

-3.49 

HLA 

October-March 

15 

24 

110 

5 

1.36 

5.69 

HBP,  HLA 

25 

14.4 

70 

5 

0.75 

3.61 

HLA 

30 

12 

40 

3 

2.88 

2.80 

LBP,  LLA 

50 

7.3 

40 

6 

0.91 

4.73 

LBP 

65 

5.5 

160 

29 

1.59 

7.05 

HLA 

Annual 

30 

12 

52 

4 

2.25 

2.16 

LBP,  LLA 

55 

6.5 

52 

8 

0.18 

1.61 

HLA 

92 

3.9 

307 

78 

2.86 

-2.13 

LBP,  LLA 

105 

3.4 

60 

17 

0.09 

2.95 

HBP,  HLA 

162 

2.2 

147 

66 

0.69 

2.15 

HLA 

360 

1 

110 

110 

1.91 

1.49 

LLA 

dominant,  whereas  on  days  of  LBP  the  30,  60,  and  90  day  cycles  were  most  sig- 
nificant. High  LA  increased  the  significance  of  the  55  and  105  day  cycles,  while 
LLA  was  associated  with  the  presence  of  30,  60,  and  90  day  cycles. 

Parameters  generated  from  the  nonlinear  analysis  (Table  II)  are  based  upon 
the  maximum  values  of  r  derived  from  the  data  based  upon  the  relative  barometric 
pressure  and  locomotor  activity.  These  values  show  slight  shifts  in  period  and 
phase  angle  (<£)  when  compared  to  the  base  spectra  (Figs.  6-9)  where  <£  was  held 
constant.  It  is  apparent  that  values  of  M  (mean  of  LA,liax  time  values)  for  each 
of  the  cycles  in  the  three  monthly  segments  varied  from  negative  to  positive  values, 
which  indicates  that  they  are  probably  oscillating  on  a  larger  period  of  about  a  year. 
This  reflects  the  greater  degree  of  LAInax  activity  prior  to  0000  hr  (negative 
values)  in  April  to  June  and  the  gradual  shift  of  activity  after  0000  hr  (positive 
values)  in  July  to  October  with  an  apparent  reversal  during  October  to  March. 


FIGURE  7.  Upper  graph  shows  a  period  analysis  of  time  of  LAma*  during  July  to  October 
indicating  by  the  solid  line  revealing  several  significant  cycles  at  5,  15,  27,  and  43  days  and  larger 
periods  ( >  90  days).  On  days  of  low  barometric  pressure  (LBP)  the  5,  27,  and  50  day 
cycles  are  dominant,  whereas  on  days  of  high  barometric  pressure  (HBP),  the  15  day  cycle 
was  the  most  significant.  Lower  graph  shows  a  similar  analysis  on  days  of  low  activity  (LLA) 
revealing  that  the  5  day  and  larger  periods  were  significant,  while  high  activity  (HLA)  was 
associated  with  15,  27,  and  50  day  cycles. 


314 


DOUGLAS    k.    ROIJERTSON 


.6 


Base 
HBP 
LBP 


.4 


.2 


(20°) 


30 


60 


90 


.6 


HLA 
LLA 


\ 


.4 


i 


30  60  90 

A  (Days)  October-March 


LOCOMOTOR  ACTIVITY  OF  FROG  315 

The  dispersion  of  activity  throughout  the  day  during  this  latter  period  is  also  re- 
flected in  the  larger  amplitude  values. 

The  cycles  that  are  dominant  in  the  annual  analysis  are  the  result  of  two  major 
fundamental  frequencies  which  possess  specific  harmonic  characteristics.  The 
first  set  of  cycles  (alpha  series)  oscillates  on  M  which  ranges  between  0.09-0.69  and 
is  represented  hy  the  55,  105,  162  day  cycles.  The  phase  angles  indicate  that  the 
55  and  105  day  cycles  are  reinforced  about  every  105  days,  and  the  162  day  cycle 
is  reinforced  hy  the  55  and  105  day  cycles  about  every  323  days.  Thus,  the  shortest 
fundamental  period  would  be  about  323  days  in  which  the  harmonics  are  6oj,  3w, 
and  2oj  with  the  first  synchronous  date  on  November  15,  1976. 

The  second  fundamental  series  of  cycles  (beta  series)  oscillates  on  M  which 
ranges  between  1.91-2.86.  The  fundamental  period  may  be  the  30  day  cycle  with 
multiples  at  w/2,  w/3,  and  oj/12  (Fig.  9),  and  the  values  of  </>  indicate  that  the  two 
significant  30  and  90  day  cycles  are  synchronized  at  days  32-34,  124,  214-216,  and 
304-308. 

By  x2  analysis  the  variables  of  barometric  pressure  and  relative  daily  activity, 
which  alter  the  significance  of  these  cycles  throughout  the  year,  have  a  significant 
(P  -  0.01)nonrandom  association  in  which  high  barometric  pressure  is  associated 
with  increased  daily  activity  and  low  pressure  with  low  daily  activity.  These  as- 
sociative features  are  also  dichotomous,  since  high  barometric  pressure  and  high 
activity  is  significantly  associated  (P  <  0.01)  with  the  alpha  cyclic  series,  while  low 
barometric  pressure  and  low  activity  is  associated  with  the  beta  cyclic  series. 

DISCUSSION 

The  spontaneous  group  activity  over  the  course  of  a  year  in  adult  male  frogs 
under  the  conditions  of  this  study  show  a  correlation  with  the  light-dark  cycle, 
ambient  barometric  pressure  and  an  exogenous  lunar  perturbation  which  appears  to 
influence  the  time  and  relative  level  of  locomotor  activity.  Each  of  these  variable 
factors  is  present  throughout  the  year,  but  the  change  in  the  L :  D  ratio  appears  to 
be  the  dominant  influence.  Absence  of  a  significant  diurnal  activity  pattern  with  a 
L:  D  ratio  of  <  1.0  indicates  that  the  length  of  time  of  the  nocturnal  period  may  be 
the  primary  influence  of  increased  nocturnal  activity,  since  the  reversed  pattern  was 
not  apparent  during  the  winter.  With  a  decreasing  dark  period,  activity  shifts 
from  primarily  nocturnal  to  auroral  at  the  time  of  the  longest  days  in  June.  Bush 
(1963)  noted  that  specimens  of  Bnfo  joivlcri  were  inactive  with  a  4L :  20D  cycle 
but  became  quite  active  as  the  L:  D  ratio  was  altered  to  12L:  12D  and  20L :  4D. 
Further,  toads  consumed  more  with  the  longer  photoperiods.  Physiological  re- 
sponsiveness of  amphibians  to  photoperiod  is  variable,  since  the  spawning  period  of 

FIGURE  8.  Upper  graph  shows  a  period  analysis  of  times  of  LA ,„..,*  from  October  to  March 
revealing  basic  cycles  (solid  line)  at  5,  15,  30,  and  50  days.  Only  the  5  day  cycle  was  altered 
significantly  under  conditions  of  high  barometric  pressure  (HBP),  while  the  50  day  cycle  was 
more  dominant  under  low  barometric  pressure  (LBP).  Lower  graph  shows  analysis  of  domi- 
nant cycles  in  periods  of  high  locomotor  activity  (HLA)  showing  significant  enhancement  of 
the  15  and  65  day  cycles,  whereas  a  30  day  cycle  was  most  significant  with  days  of  low  locomo- 
tor activity  (LLA). 


316 


DOUCI.AS   R.   kOI'.KKTSnX 


]\\ina  temporaria  appears  to  be  controlled  by  temperature  and  not  by  light  (van 
Oordt  and  van  Oordt,  1955)  ;  but  the  spermatogenic  cycle  in  the  salamander 
PIctJiodon  cincreus  is  primarily  regulated  by  photoperiod  (Werner,  1969).  The 


.4  I — 


r.2 


(52°) 


50 


100 


200 


300 


r.2  — 


/  \                                                 HLM 

/        '                                                                        1   I    A 

f 

(l                                                                       L  L  A 

i           1 
.            I 
\ 
/              \ 
1               1 

r         i 
/  •        l 

/'  -:    :.        i                 /  —  x 

1       :'        '.             \                         f             \ 

1       \     I        / 

,'     i        ':         ' 

/  >       i         \     \       '            N                                  ,'' 

\    \\      \ 

?  ;-;  MV;      \      /  ./ 

;i   /  \                              \           / 

- 

i  l          •          :           :                                                                                     »                                '      ..-• 

1 

1 
VJ 

\  ."••"      / 

\                                                             .!•'                    / 

:      /                              ':                                                       .••''  l                    / 

1               '                                              -•''       1      '               '                                                 1 
1               -                  1                '                    1                      .             1      '.            '                1 

50                  100                                       200                                     30C 

A  (Days) 


Annual 


FIGURE  9.  Upper  graph  shows  a  period  analysis  of  all  LAmax  data  collected  throughout  the 
year  revealing  major  cycles  at  30,  55,  105,  and  160  days.  Under  conditions  of  lower  than  the 
yearly  mean  barometric  pressure  (LBP),  significant  cycles  were  apparent  at  30,  60,  90,  and  155 
days,  whereas  with  higher  than  the  yearly  mean  barometric  pressure  (HBP)  the  dominant 
cycles  were  at  50,  105,  and  175  days.  Lower  graph  shows  a  similar  analysis  of  those  cycles 
most  dominant  when  frogs  were  least  active  (LLA),  present  at  30,  60,  and  90  days,  while 
cycles  apparent  on  days  of  high  activity  (HLA)  were  at  50,  105,  and  165  days. 


LOCOMOTOR  ACTIVITY  OF  FROG  317 

increase  in  nocturnal  activity  with  the  beginning  of  longer  <lavs  may  be  an  indepen- 
dent but  synchonized  part  of  the  initiation  of  sexual  activity  after  the  spring  equinox. 

The  influence  of  temperature  on  amphibian  behavioral  activity  has  been  noted 
by  several  authors.  Early  observations  by  Torelle  (1903)  of  specimens  of  Rana 
viresccns  viresccns  (Rana  pipiens  pipiens*)  found  that  at  temperatures  :  10°  C, 
frogs  moved  away  from  light;  whereas  with  temperatures  >  30°  C  frogs  were  at- 
tracted to  light.  Higginbotham  (1939)  observed  that  with  a  10°  C  rise  in  tempera- 
ture there  was  a  doubling  to  tripling  of  the  amount  of  activity  of  toads  (Bnfo 
amcricanus  and  Bitfo  foivlcri )  but  no  alteration  in  the  overall  pattern  of  activity. 
Bujo  jou'leri  also  appears  to  be  most  active  between  19.6—24.7°  C  in  April  to  May 
(Martof,  1962)  and  consumes  twice  as  much  food  with  a  10°  temperature  rise 
from  21°  C  to  31°  C  (Bush,  1963).  These  responses  in  behavioral  activity  with 
change  in  ambient  temperature  is  of  importance  in  the  interpretation  of  the  present 
data.  Over  the  course  of  a  year  the  temperature  range  was  held  at  8°  C  and  was 
within  the  limits  observed  by  Martof  (1962)  where  toads  are  most  active.  How- 
ever, within  the  temperature  range  of  the  present  study  there  was  a  nonsignificant, 
and  somewhat  negative  relationship,  between  total  daily  activity  and  the  prevailing- 
temperature.  Within  any  monthly  period,  the  maximal  temperature  range  wras  only 
3°  C,  which  was  considerably  less  than  the  10°  C  differential  associated  with  varia- 
tions in  total  activity  of  toads  (Higginbotham,  1939;  Bush,  1963).  Thus,  while  it 
is  possible  that  the  slight  variations  in  temperature  could  modify  portions  of  the 
observed  activity,  clear  relationships  were  not  as  apparent  as  those  associated  with 
the  light-dark  cycle,  lunar  frequencies  and  barometric  pressure. 

The  exogenous  lunar  influence  which  has  been  described  by  others  (Ralph. 
1957;  FitzGerald  and  Bider,  1974)  to  affect  amphibian  activity  patterns  was  also 
apparent  in  the  present  study.  Time  of  high  transit  was  associated  with  an  en- 
hancing and  depressing  effect  during  daylight  hours  in  winter  and  spring  to  early 
summer,  respectively,  and  represents  a  possible  mechanism  to  induce  oscillatory 
patterns  in  activity  as  seen  in  the  activity  isogram  for  April  to  June.  The  patterns 
may  be  modified  with  the  varying  length  of  daylight  hours.  The  fact  that  the 
activity  isogram  in  the  fall  is  not  a  replica  of  the  pattern  observed  in  the  spring  may 
reflect  the  reversal  in  activity  levels  of  these  animals  during  daylight  to  this  exo- 
genous influence. 

The  cyclic  patterns  of  locomotor  activity,  as  revealed  by  wave  analysis,  in  the 
three  major  monthly  periods  and  for  the  entire  year  appears  to  represent  a  con- 
sistent phenomenon.  That  is,  the  time  of  maximal  activity  each  day  occurred  at  a 
similar  time  period  at  a  later  date  which  was  determined  by  the  dominant  cycle 
present.  This  is  best  exemplified  by  the  lunar  cycle  which  approximates  30  days 
(beta-series)  and  is  apparent  in  the  isogram  of  locomotor  activity  for  April  to  June. 
But  in  addition  to  this  more  obvious  lunar  cycle,  there  are  other  cycles  that  are 
generated  by  lunar  and  solar  movement  which  can  give  rise  to  tidal  harmonics 
(Godin,  1972).  The  principal  lunar  constituent  (M^)  has  a  period  of  12.42  hours 
and  other  cycles  "beat"  against  Mo  giving  rise  to  a  lunar  fortnightly  tide  (13.66 
days),  a  monthly  modulation  of  27.57  days  and  a  solar  semiannual  tide  of  182.9 
days.  Thirty-four  separate  tidal  frequencies  have  been  identified  (Godin,  1972), 


318  DOUGLAS  R.  ROBERTSON 

rnch  of  which  can  generate  their  own  harmonics  which  can  reinforce  or  dampen 
one  another.  The  activity  cycles  which  are  observed  in  frogs  may  lie  affected  by 
such  tidal  harmonics,  the  relative  dominance  of  each  reflected  in  the  value  of  r  at 
the  optimal  phase  angle.  For  example,  the  characteristics  of  the  alpha-series  of 
cycles,  with  a  fundamental  period  of  323  days,  suggests  that  it  may  be  related  to 
the  eclipse  year  of  346.6  days  (a  "beat"  frequency  of  the  synodic  and  clraconitic 
months),  since  the  first  synchronous  date  of  all  harmonics  was  nine  days  after  a 
lunar  eclipse  on  11/6-7/76.  The  next  ideal  synchronous  date  is  10/5/77  which  is 
between  two  eclipses  on  9/27/77  (lunar)  and  10/12/77  (solar)  (American  Ephcm- 
cris  and  Nautical  Almanac,  1976,  1977).  Confirmation  of  such  a  relationship 
would  require  analysis  of  locomotor  activity  over  a  period  of  several  eclipse  years. 

Of  interest  is  that  these  cycles  are  present  from  October  to  March,  a  period 
characterized  by  continuous  24  hour  activity  in  the  absence  of  a  defined  diurnal 
pattern.  Period  analysis  shows  that  they  are  quite  similar  to  the  cycles  observed 
in  April  through  June  and  thus  appear  to  be  independent  of  the  light-dark  cycle 
and  may  be  similar  to  the  lunar  perturbations  observed  by  Ralph  (1957)  in  the 
salamander  kept  in  continuous  darkness.  These  oscillations  may  be  expressed  as 
overt  activity  patterns  when  the  light-dark  effect  is  significant  to  induce  a  defined 
nocturnal  pattern.  While  it  may  be  argued  that  these  are  statistical  cycles  and  not 
overt,  patterns  can  be  identified  in  composite  isograms.  Further,  they  can  be 
mathematically  described  as  a  first  order  approximation  to  a  sine  wave  which  al- 
lows a  possible  means  of  prediction  (Table  II). 

An  additional  feature  of  the  wave  analysis  is  that  other  variables  such  as  baro- 
metric pressure  and  relative  levels  of  locomotor  activity  can  be  related  to  two  funda- 
mental cycles  which  might  otherwise  be  difficult  to  detect.  Brown  ct  <<?/.,  (1955) 
noted  that  oxygen  consumption  in  the  salamander  Triturus  was  inversely  related  to 
barometric  pressure  but  not  sharply  defined.  Even  in  the  present  study  a  sig- 
nificant inverse  relationship  can  be  derived  between  absolute  daily  lomomotor  activ- 
ity and  absolute  barometric  pressure  during  May  (r  -  —0.0418:  P  —  0.02),  but  it 
is  not  consistent  at  other  months.  The  present  analysis  reduces  the  apparent 
"biological  noise"  by  defining  specific  days  and  times  of  LAimix  which  can  be  cor- 
related with  prevailing  barometric  pressure. 

The  sensitivity  of  animals  to  atmospheric  pressure  changes  has  been  observed  in 
aquatic  invertebrates  such  as  Ca  rein  us  (Nay  lor  and  Atkinson,  1972)  and  to  hydro- 
static pressure  as  in  the  amphibian  tadpoles,  Rana  sih'atica  and  Amblystoma  (John- 
son and  Flagler,  1951  ),  which  in  both  cases  increase  their  activity  with  an  increase 
in  pressure.  Conversely,  in  a  terrestrial  form  such  as  a  newly  hatched  chick,  in- 
creased activity  is  associated  with  a  decrease  in  barometric  pressure  (Bateson, 
1974). 

In  the  present  study  this  apparent  correlation  of  barometric  pressure  to  activity 
level  was  specific  for  certain  cycles.  The  relationship  of  these  two  variables  is 
unclear,  but  lunar  and  solar  movement  also  generates  atmospheric  tides  in  addition 
to  marine  tides  (Siebert,  1961).  The  lunar  component  can  induce  a  barometric 
variation  of  0.001  millibars  (0.025  mm  Hg),  while  the  solar  component  can  induce 
a  variation  up  to  0.03  millibars  (0.76  mm  Hg).  Detection  of  such  variations  in 


LOCOMOTOR  ACTIVITY  OF  FROG  319 

barometric  pressure  exclusive  of  the  daily  weather  perturbations  might  induce  the 
observed  cyclic  patterns.  In  view  of  the  dichotomy  of  the  alpha  and  beta  series  of 
cycles  which  are  associated  with  specific  variables,  the  resultant  patterns  of  activity 
may  reflect  two  independent  receptor-effector  behavioral  systems. 

In  summary,  the  daily  and  annual  solar  light  pattern  appears  to  be  related  to 
the  diurnal  and  nocturnal  activity  of  the  adult  male  frog,  Rana  pif>icns,  but  it  has 
activity  patterns  of  specific  periodicities  which  are  similar  to  lunar  periodicities. 
Additionally,  there  are  significant  correlations  between  these  cyclic  patterns  with 
the  prevailing  barometric  pressure  and  the  level  of  locomotor  activity.  The  general 
behavior  patterns  of  this  amphibian  over  the  course  of  a  year  exhibits  several  cor- 
relates to  basic  geophysical  forces.  The  significance  of  these  relationships  may  be 
reflected  in  similar  activity  patterns  of  organisms,  such  as  insects  (see  reviews  of 
Harker,  1958;  Corbet,  1960),  which  enter  into  the  amphibian  food  chain.  Such 
synchrony  could  increase  the  probability  of  successful  feeding  and  ultimate  survival. 
Variations  in  intestinal  transport  activity  (Robertson,  1976)  may  reflect  the  gen- 
eral physiological  cycling  coherent  with  these  food  gathering  processes. 


The  major  portion  of  this  project  was  supported  by  a  grant  from  the  National 
Science  Foundation,  No.  BMS74- 19330.  The  technical  assistance  of  Mrs.  Nancy 
Cheever  is  gratefully  acknowledged.  I  also  wish  to  thank  Mr.  Edward  Matyas, 
Computer  Services  of  Upstate  Medical  Center,  for  construction  of  the  FORTRAN  iv 
Program. 

SUMMARY 

The  spontaneous  locomotor  activity  (LA)  of  adult  male  frogs  (Rana  pipicns, 
Northern  variety)  was  monitored  throughout  the  year  in  an  apparatus  which 
detected  vertical  water  movements.  Frogs  exposed  to  the  seasonal  change  in 
ambient  light  and  maintained  at  a  constant  mean  annual  temperature  of  19.3  ± 
3.1°  C  exhibited  significant  correlations  of  activity  to  the  light-dark  cycle,  baro- 
metric pressure  and  lunar  perturbations.  AYhen  the  light :  dark  ratio  was  <  1.0 
(October  to  March)  frogs  displayed  "random"  activity  throughout  the  24  hr 
period;  but  with  the  L:D  ratio  between  1.0-1.45  activity  was  primarily  nocturnal 
between  2100-0000  hr  and  at  sunrise,  while  with  a  L:D  Ratio  >  1.45  maximal 
activity  occurred  at  sunrise.  Activity  also  was  correlated  with  time  of  lunar  high 
transit  (HT)  where  occurrence  of  HT  during  daylight  hours  in  April  to  June  was 
associated  with  depressed  activity,  while  HT  during  daylight  in  October  to  March 
was  coincident  with  elevated  period  of  activity.  Use  of  a  FORTRAN  iv  Program  to 
analyze  time  of  maximal  LA  each  day  throughout  the  year  revealed  oscillatory  be- 
havior patterns  with  periods  similar  to  lunar  tidal  cycles.  An  alpha-series  of  cycles 
(55.  105,  and  162  day  periods)  were  significantly  associated  and  dominant  on  days 
of  high  barometric  pressure  (above  the  annual  mean  of  761  mm  Hg)  and  char- 
acterized by  high  levels  of  activity  (above  the  annual  mean  of  313  events/day).  A 
beta-series  (30,  60,  90  day  cycles)  was  dominant  on  days  of  low  barometric  pres- 


320  DOUGLAS  R.  ROBERTSON 

sure  (<  761  mm  Hg)  and  coincident  with  low  levels  of  activity  (<  313  events/ 
day).  Spontaneous  activity  of  frogs  apparently  is  not  random,  but  reflects  an  as- 
sociation with  basic  geophysical  forces  which  elicit  a  complex  but  definable  behavior 
pattern. 

LITERATURE  CITED 

BATESON,  P.  P.  G.,  1974.     Atmospheric  pressure  during  incubation  and  post-hatch  behavior  in 

chicks.    Nature,  248  :  805-807. 
BROWN,  F.  A.,  H.  M.  WEBB,   AND  E.  J.   MACEY,   1957.     Lag-lead  correlations  of  barometric 

pressure  and  biological  activity.     Biol.  Bull.,  113:   112-119. 
BROWN,  F.  A.,  JR.,  H.  M.  WEBB,  M.  F.  BENNETT,  AND  M.  I.  SANDERS,  1955.     Evidence  for  an 

exogeneous   contribution   to   persistent   diurnal   and    lunar   rhythmicity   under   so-called 

constant  conditions.     Biol.  Bull..  109:  238-254. 
BUSH,   F.  M.,   1963.     Effects  of  light  and  temperature  on  the  gross  composition  of  the  toad 

Bufo  fozvlcri.   J.  Exp.  Zool.,  153:  1-13. 
CHUGUNOV,  Y.   D.,  AND  E.   V.   KUZNETSOV,   1972.     Asynchronous  daily  rhythms  of  locomotor 

activity  and  oxygen  consumption  in  Rana  temporaria.    Zool.  Z/z.,  50:  1695-1699. 
CORBET,   P.   S.,   1960.     Patterns  of  circadian   rhythms   in   insects.     Cold  Spring  Harbor  Symp. 

Quant.  Biol..  25  :  357-360. 
DOLE,  J.   W.,   1972.     Evidence  of  celestial   orientation   in   newly-metamorphosed   Rana   pipiens. 

Herpetologica,  28 :  273-276. 
FERGUSON,  D.  E.,  AND  H.  F.   LANDRETH,   1966.     Celestial  orientation  of  Fowler's  Toad  Bufo 

fozvleri.  Behavior,  26:  105-123. 
FERGUSON,  D.  E.,  H.  F.  LANDRETH,  AND  M.  R.  TURNIPSEED,  1965.     Astronomical  orientation  of 

the  southern  cricket  frog.  Acris  gryllus.    Copcia.  1965  :  58-66. 
FITZGERALD,  G.  J.,  AND  J.  R.   BIDER,   1974.     Influence  of  moon  phase  and  weather  factors  on 

locomotor  activity  in  Bufo  aincricanus.    Oikos,  25:  338-340. 

GIBBONS,  J.  W.,  AND  D.  H.  BENNETT,   1974.     Determination  of  anuran  terrestrial  activity  pat- 
terns by  a  drift  fence  method.     Copcia,  1974 :  236-243. 

GODIN,  G.,  1972.     The  analysis  of  tides.    University  of  Toronto  Press,  Toronto,  653  pp. 
HIGGINBOTHAM,   A.    C.,    1939.      Studies   on   amphibian   activity — I.    Preliminary   report   on   the 

rhythmic  activity  of  Bufo  aincricanus  Holbrook  and  Bufo  fowlcri  Hinkley.    Ecology, 

20 :  58-70. 

MARKER,  J.,  1958.     Diurnal  rhythms  in  the  Animal  Kingdom.  Biol.  Rev.,  33  :   1-52. 
JAEGER,  G.,  J.  P.  HAITMAN,  AND  L.  S.  JAEGER,   1976.     Bimodal  diel  activity  of  a  Panamanian 

dendrobatid  frog,  Colostethus  nubicola,  in  relation  to  light.   Herpetologica,  32:  77-81. 
JOHNSON,  F.  H.,   AND  E.  A.  FLAGLER,   1951.     Activity  of  narcotized  amphibian   larvae   under 

hydrostatic  pressure.    /.  Cell  Comp.  Physio!.,  31  :  15-25. 
KUZNETSOV,  E.  V.,  Y.  D.  CHUGUNOV,  AND  V.  Y.  BRODSKII,  1972.     Biological  daily  rhythms  in 

frogs.  Rana  csculcnta  under  natural  conditions.     Z/;.  Obsoch.  Biol.,  33  :  210-216. 
LANDRETH,  H.   F.,  AND  D.  E.   FERGUSON,   1966.     Evidence  of  suncompass   orientation   in  the 

chorus  frog,  Pscudacris  triscriata.    Herpetologica,  22:  106-112. 
LANDRETH,  H.  F.,  AND  D.  E.  FERGUSON,  1967.     Newts :  suncompass  orientation.     Science,  158: 

1459-1461. 
LINZEY,  D.  W.,  1967.     Food  of  the  leopard  frog,  Rana  pipiens  pipifiis  in  central   New  York. 

Herpetologica,  23:  11-17. 
MARTOF,   B.   S.,   1962.     The  behavior  of  Fowler's  toad   under   various   conditions  of  light  and 

temperature.    Pliysiol.  Zool.,  35  :  38-46. 
NAYLOR,  E.,  AND  R.  J.  A.  ATKINSON,  1972.     Pressure  and  rhythmic  behavior  of  inshore  marine 

animals.   Syinp.  Soc.  Exp.  Biol.,  26:  395-415. 

OLDROYD,  H.,  1960.     Insects  and  their  ivorld.     University  of  Chicago  Press,  Chicago,  139  pp. 
PARK,  O.,  A.  BARDEN,  AND  E.  WILLIAMS,  1940.     Studies  in  nocturnal  ecology.  IV.     Further 

analysis  of  activity  in  Panama  rain  forest  animals.     Ecology,  21  :   122-134. 
RALPH,  C.  L.,  1957.     A  diurnal  activity  rhythm  in  Plethodon  cinereus  and  its  modification  by 

an  influence  having  a  lunar  frequency.    Biol.  Bull.,  113:  188-197. 


LOCOMOTOR  ACTIVITY  OF  FROG  321 

ROBERTSON,   D.   R.,    1976.     Diurnal   and   lunar   periodicity   of   intestinal   calcium   transport   and 

plasma  calcium  in  the  frog,  Rana  pipiois.    Comp.  Biochcm.  Physiol.,  54A :  225-231. 
SIEBERT,  M.,  1961.    Atmospheric  tides.   Advanc.  Geophysics.  7:  105-187. 
SKAIFE,  S.  H.,  1961.     The  study  of  ants.    Spottiswoode,  Ballantyne  and  Co.,  Ltd.,  London,  177 

pp. 

TORELLE,  E.,  1903.    The  response  of  the  frog  to  light.  Am.  J.  Physio!.,  9 :  466-488. 
VAN  OORDT,  G.  J.,  AND  P.  G.  W.  J.  VAN  OORDT,  1955.    The  regulation  of  spermatogenesis  in  the 

frog.    Mem.  Soc.  EndocrinoL,  4 :  25-38. 
VAN  BERGEITK,  W.  H.,  1967.    Anticipatory  feeding  behavior  in  the  bullfrog  (Rana  catesbciana) . 

Anim.  Behav.,  15:  231-238. 
WERNER,  J.  K.,   1969.     Temperature-photoperiod  effects  on  spermatogenesis  in  the  salamander 

Plethodon  cinereus.    Cofeia,  1969:  592-602. 
WHITAKER,  J.  O.,  JR.,  1961.     Habitat  and  food  of  mousetrapped  young  Rana  pipicns  and  Rana 

clamitans.   Hcrpetologica,  17:  173-179. 
WRIGHT,   A.   H.,   AND  A.  A.   WRIGHT,    1949.     Handbook   of  froys  and   toads   of   the    United 

States  and  Canada.    Cornell  University  Press,  Ithaca,  New  York,  640  pp. 


:    li'wl.  Hull..  154 :   .U2-3.54.    (April,  1978) 


SEASONAL  RESPIRATION  IN  THE  MARSH  PERIWINKLE, 

LITTOKINA  IRRORATA 

THOMAS  C.  SHIRLEY,  GUY  J.  DENOUX  1  AND  WILLIAM  B.  STICKLE 

Dcptii'hnt'iil  of  Zooloi/y  and  Pliysinlin/y,  Louisiana  State  I'nircrsity, 
Baton  RIIHI/C,  Louisiana  70X03  U.S.A. 

Intertidal  invertebrates  inhabit  an  environment  which  varies  on  a  diurnal,  tidal 
and  seasonal  basis.  It  is  not  surprising  that  the  relationship  between  their  meta- 
bolism and  the  environment  may  be  complex.  Adaptations  to  the  intertidal  en- 
vironment by  some  invertebrates  include  a  standard  metabolic  rate  that  is  insen- 
sitive to  a  wide  range  of  temperatures  and  an  active  metabolic  rate  that  is  tem- 
perature-independent, or  thermo-neutral,  within  a  zone  centered  about  the  ambient 
field  or  acclimation  temperature  (summarized  by  Newell,  1969,  1973,  1976).  The 
relationship  between  metabolism  and  environmental  variables  may  be  obscured  by 
other  adaptations,  such  as  the  presence  of  diurnal  or  tidal  metabolic  rhythms 
(Sandeen,  Stephens  and  Brown,  1954;  Sandison,  1966;  Shirley  and  Findley, 
1978).  Seasonal  acclimatization  and  acclimation  temperature  may  also  greatly 
affect  metabolic  rate  and  range  of  the  temperature-insensitive  zone  ( Pye  and  Newell, 
1973).  Many  metabolic  studies  of  invertebrates  have  failed  to  consider  these  pos- 
sible complexities  of  invertebrate  metabolism  or  have  not  been  of  a  long  enough 
duration  to  investigate  seasonal  changes. 

The  marsh  periwinkle,  Littorhni  irrorata  (Say,  1822),  is  widely  distributed 
from  New  York  to  Texas  (Bequaert,  1943)  and  is  the  most  important  gastropod 
in  terms  of  biomass  in  the  salt  marshes  of  the  Gulf  of  Mexico  (Day,  Smith,  Wag- 
ner, and  Stowe,  1973;  Subrahmanyam,  Kruczynski  and  Drake,  1976;  Hamilton, 
1976).  Its  well-studied  life  history  (Bingham.  1972a,  b;  Alexander,  1976;  Ocluni 
and  Smalley,  1959;  Shirley  and  Findley,  1978)  and  the  concurrent  investigation  of 
its  biochemical  composition  and  body  component  indexes  (Bistransin,  1976)  are 
useful  in  understanding  the  snail's  metabolic  patterns.  The  snail  is  supratidal  in 
habit,  normally  found  on  Spartina  stems  above  the  air-water  interface.  Although 
a  number  of  respiration  studies  of  intertidal  gastropods  have  been  reported  (Bert- 
ness  and  Schneider,  1976;  Coleman.  1976;  Huebner,  1973;  McMahon  and  Russell- 
Hunter,  1977;  Newell  and  Pye,  1970a.  b,  1971;  Sandison,  1966,  1967).  relatively 
few  have  been  performed  with  supratidal  marine  snails.  A  trait  of  L.  irrorata  also 
meriting  investigation  is  its  ability  to  attach  to  a  stem  by  means  of  a  mucous  hold- 
fast and  withdraw  into  its  shell,  supposedly  to  avoid  unfavorable  conditions.  The 
effect  of  formation  of  mucous  holdfasts  on  the  metabolism  of  the  snail  has  not 
previously  been  examined. 

The  investigation,  therefore,  focused  on  three  main  areas :  first,  the  possibility 
of  diurnal  metabolic  rate  rhythms ;  secondly,  seasonal  changes  in  metabolic  rate- 

1  Present  address :  Department  of  Oceanography,  Texas  A&M  University,  College  Station, 
Texas  77843  U.S.A. 

322 


RESPIRATION  IN  LITTORINA  IRRORATA  323 

temperature  curves;  and  thirdly,  the  metabolic  rate  of  snails  in  conditions  which  are 
conducive  to  mucous  holdfast  formation. 

MATERIALS  AND  METHODS 

Snails  were  collected  at  monthly  intervals  for  15  months  beginning  in  October, 
1973,  from  a  Spartina  altcrniflora  salt  marsh  located  3.5  km  northwest  of  Grande 
Isle,  Louisiana.  The  snails  were  transported  to  the  laboratory  and  maintained 
under  salinity,  temperature  and  photoperiodic  regimen  corresponding  to  measured 
field  conditions  (Table  I).  Meteorological  data  were  obtained  from  the  U.  S. 
Coast  Guard  Station  on  Grande  Isle  in  order  to  consider  the  effects  of  acclimatiza- 
tion temperatures  on  experimental  results.  Inasmuch  as  the  snails  are  supratidal 
in  habit,  air  temperature  was  considered  to  be  of  greatest  importance  and  tempera- 
ture means  were  determined  from  measurements  made  at  three  hour  intervals  for 
the  duration  of  the  study.  Studies  were  initiated  on  the  day  that  snails  were  collected 
and  respiration  measurements  of  the  snails  were  usually  completed  within  three  to 
four  days  after  collection.  Individual  oxygen  consumption  rates  of  14  snails  were 
measured  for  each  uptake  determination  with  a  Gilson  Differential  Respirometer 
using  standard  manometric  techniques.  The  largest  snails  present  in  the  field  (x  = 
152  mg  dry  tissue  wt)  were  selected  for  all  determinations.  The  110  ml  reaction 
vessels  and  snails  were  allowed  to  equilibrate  for  one  hour  prior  to  the  initiation 
of  measurements.  Measurements  were  made  for  approximately  one  hour  and  cor- 
rected to  STP.  The  reaction  vessels  were  not  shaken  during  the  experiment.  A 
paper  wick  and  0.5  ml  of  30%  KOH  were  placed  in  the  sidearm  of  each  vessel. 

Three  separate  studies,  each  with  different  experimental  procedures,  were  con- 
ducted. Studies  1  and  2  were  conducted  under  water  saturated  conditions.  Eight 
milliliters  of  artificial  sea  water  (Instant  Ocean)  of  the  same  salinity  as  in  the  field 
at  the  time  of  collection  were  placed  in  each  reaction  vessel.  In  study  1  the  respira- 
tion of  14  snails  was  measured  for  a  one  hour  period  on  alternate  hours  for  36 
hours  at  a  5°  C  temperature  increment  ±2.5°  C  of  the  field  temperature  at  the  time 
of  collection.  Study  2  consisted  of  a  one  hour  measurement  of  respiration  of  snails 
during  daylight  hours  at  5°  C  temperature  increments  from  5  to  45°  C.  Fourteen 
different  snails  were  used  for  each  temperature.  In  study  3  an  attempt  was  made 
to  induce  the  snails  to  form  mucous  holdfasts  by  subjecting  them  to  low  humidity 
conditions.  Accordingly,  no  water  was  added  to  the  vessels  and  additionally, 
mantle  compartment  water  was  removed  by  gently  pushing  the  snail  as  far  as  pos- 
sible into  its  shell  after  which  the  snail  was  dried.  A  drying  tube  filled  with  calcium 
sulfate  was  attached  to  the  air  intake  of  the  respirometer,  and  the  snails  were  ex- 
posed to  the  dry  air  overnight  prior  to  respiration  measurements.  Measurements 
in  study  3  were  made  at  the  same  temperature  as  study  1 . 

Observations  on  the  condition  of  the  snails  and  formation  of  mucous  holdfasts 
were  noted  after  each  respiration  measurement.  After  completion  of  each  experi- 
ment, the  shell  of  each  snail  was  cracked  and  the  soft  parts  removed  carefully  In- 
hand  and  dried  to  constant  weight  at  85°  C. 

Regression  analyses  of  login  oxygen  consumption  per  animal  against  Iog10  dry 
tissue  weight  were  determined  for  each  experiment  by  the  least  squares  method. 


324 


SHIRLEY,  DENOUX,  AND   STICKLE 


Inasmuch  as  the  majority  of  regression  equations  were  not  significant  (P  >  0.05), 
mean  weight  specific  oxygen  consumption  rates  [/zl  O2/(g-hr)]  and  confidence  in- 
tervals at  the  95%  level  were  calculated  for  all  experiments.  In  study  1,  analysis  of 
variance  of  the  randomized  hlock  design  was  performed  to  determine  differences  in 
respiration  rates  with  respect  to  time  of  day.  Further  partitioning  of  the  variance 
was  tested  by  orthogonal  comparisons.  In  study  2,  Qi0  values  over  5°  C  tempera- 
ture intervals  were  determined,  and  the  significance  of  each  was  tested  against  the 
null  hypothesis  that  Qio  =:  1.0  using  a  modified  Z-test  (Snedecor  and  Cochran, 
1971).  In  study  3,  regression  equations,  mean  respiration  rates  and  95%  con- 
fidence intervals  were  calculated  separately  for  those  snails  that  had  formed  mucous 
holdfast  and  those  that  had  not. 

One  of  the  principal  modifying  agents  of  metabolism  is  activity,  and  several 
methods  of  coping  with  animal  activity  have  been  attempted  to  reduce  scatter  about 
the  regression  lines  relating  log  metabolism  to  log  body  weight.  One  of  the  more 
recent  methods  used  by  Newell  and  associates  (Newell  and  Northcroft,  1965; 
Newell  and  Pye,  1971  ;  Newell  and  Roy,  1973)  has  been  to  separate  active  rates 
from  standard  rates  solely  on  the  basis  of  magnitude  without  correlation  to  animal 
activity.  Although  this  method  has  been  used  with  success  in  significantly  reducing 
regression  scatter,  it  ignores  the  activity-metabolism  and  individual  variation 
(Barnes  and  Barnes,  1969;  Coleman,  1976).  The  concept  of  metabolism  increas- 
ing directly  with  activity  is  obvious.  Difficulties  in  correlating  measured  activity 
with  metabolism  have,  however,  been  encountered  by  several  investigators  (McFar- 
land  and  Pickens,  1965;  McLusky,  1973).  Additional  complications  in  L.  irrorata 


TABLE  I 


Average  rale  of  oxygen  consumption  (Qo-2)  of  Littorina  irrorata  over  24  hr,  expressed  as  pi  O-if  (g  dry 
wt-hr);  ±V5'lo  confidence  interval  (C.I.) 


Month 

Temperature  °  C 

Orthogonal 

day/night 

A"  dry  weight 
of  snails 
(mg) 

Oxygen  consumption 

Field 

Expt. 

Dav 
A"  ±95%  C.I. 

Night 
A'  ±  95%  C.I. 

Oct.  73 

24 

25 

** 

121 

540.6  ±  35.0 

615.0  ±  25.4 

Dec.  73 

19 

15 

** 

148 

199.5  ±  15.0 

242.3  ±  13.6 

Jan.  74 

12 

15 

NS 

160 

245.4  ±  12.9 

241.8  ±  10.9 

Feb.  74 

19 

25 

** 

169 

530.2  ±  29.3 

622.3  ±  34.2 

Mar.  74 

19 

25 

** 

201 

561.1  ±  19.4 

673.5  d=  37.5 

Apr.  74 

20 

25 

NS 

193 

586.6  ±  28.3 

589.4  ±  38.9 

May  74 

23 

30 

** 

203 

649.4  ±  20.0 

708.3  ±  27.1 

Jun.  74 

27 

30 

** 

173 

822.2  ±  37.4 

1039.2  ±  75.8 

Jul.  74 

26 

30 

** 

163 

893.7  ±  63.3 

968.7  ±  70.2 

Aug.  74 

27 

30 

** 

175 

735.0  ±  28.7 

923.3  ±  45.5 

Sep.  74 

26 

30 

** 

121 

824.3  ±  34.5 

921.5  ±  47.4 

Oct.  74 

21 

30 

** 

127 

542.4  ±  26.9 

636.3  ±  34.5 

Nov.  74 

19 

15 

NS 

144 

236.4  ±  16.0 

220.2  ±  11.4 

Dec.  74 

13 

15 

** 

148 

238.0  ±  11.5 

264.9  ±  15.1 

*  P  <  0.01. 
NS  =  P  >  0.05. 


RESPIRATION  IX  L1TTORINA  IRRORATA  325 

are  the  snail's  positive  phototropism  (Bingham,  1972a)  and  a  lower  metabolic  rate 
during  light  than  dark  (Shirley  and  Findley,  1978).  Illumination  of  flasks  to 
monitor  activity  was  therefore  avoided  during  dark  conditions  in  this  study.  Move- 
ment of  reaction  flasks  containing  specimens  was  also  avoided  to  preclude  disturbing 
the  specimens  and  increasing  metabolism  (Newell,  Weiser  and  Pye,  1974;  Aldrich, 
1975).  Since  the  snails'  activities  were  not  measured,  all  data  were  utilized  in  re- 
gression analyses  and  computations  without  arbitrarily  assigning  a  basal  or  active 
rate.  The  term  "basal"  metabolic  rate  has  a  number  of  specific  criteria  which 
cannot  be  readily  applied  to  poikilothermic  invertebrates,  especially  gastropods,  be- 
cause of  the  plasticity  of  their  oxygen  consumption  (Lewis,  1971 ;  McMahon  and 
Russell-Hunter,  1977;  Russell-Hunter,  1964;  Sandison,  1967).  The  data  are 
hopefully  indicative  of  natural  metabolic  patterns  in  study  1  and  indicative  of 
routine  or  normally  active  snails  in  the  respiration-rate  temperature  study.  Mc- 
Mahon and  Russell-Hunter  (1977)  used  a  similar  approach  in  their  work  with 
littoral  snails. 

The  percentage  of  caloric  content  respired  per  day  per  snail  was  determined  by 
month  for  the  year  1974.  The  total  volume  of  oxygen  consumed  per  gram  dry 
weight  of  snail  per  day  at  the  average  air  temperature  of  each  month  was  calcu- 
lated, with  adjustment  for  the  increased  night  consumption  for  appropriate  months. 
Oxygen  consumption  was  converted  to  caloric  values  by  means  of  an  oxycaloric 
coefficient  of  4.8  cal/ml  O2  (Crisp,  1971).  The  oxycaloric  value  was  adjusted  to 
that  of  the  mean  weight  of  snails  for  each  month.  The  total  dry  weight  of  carbo- 
hydrate, lipid  and  protein  per  snail  for  each  month  from  the  same  population 
sample  used  for  the  respiration  investigation  was  obtained  from  the  work  of  Bis- 
transin  (1976)  and  used  to  determine  the  total  caloric  value  for  the  mean  weight 
snail  per  month.  Division  of  the  caloric  value  of  respiration  per  day  by  the  total 
caloric  content  resulted  in  the  percentage  of  total  calories  respired  per  day. 

RESULTS 

One  of  the  principal  parameters  which  might  be  expected  to  influence  meta- 
bolism is  the  acclimatization  temperature.  The  average  air  temperature  for  each 
month,  determined  from  measurements  made  at  3  hr  intervals  for  the  entire  year, 
is  listed  on  Table  I. 

Regression  analyses  of  log]0  oxygen  consumption  versus  Iog10  dry  tissue  weight 
were  significant  (P  <  0.05)  in  only  72  of  256  regression  equations  in  study  1. 
Oxygen  consumption,  expressed  as  /*!  O2/(g  dry  wt-hr)  (Table  I)  was  therefore 
not  normalized  from  regression  equations,  but  rather  adjusted  to  per  gram  dry 
tissue  wt.  No  pattern  was  found  in  the  occurrence  of  significant  regression  equations 
according  to  time  of  day  or  year.  The  nine  months  with  weight  ranges  of  snails 
greater  than  67  mg,  with  two  exceptions,  had  the  greatest  number  of  significant  re- 
gression equations. 

Highly  significant  differences  in  respiration  rates  with  respect  to  time  of  day 
were  found  by  ANOVA  in  all  months  except  February,  1974.  Further  partitioning 
of  variance  by  orthogonal  analysis  demonstrated  that  highly  significant  increased 


326 


SIllkl.KY,    DKXOUX.   AND   STICKLE 


> 


•ti 


2 
s 

1 

? 

o 
Cv 


— <  O  «-<  O  oo  i- 

-t    01  — I 


LO  *^1   ""}    -^M   O   »O 

^— i  GO   r/}   »— *  ON 


>O 

OO 


r<3    i—  l   CN   CN 


rf)   O   O   -t<   "3    ^H   vO   'O 


OS   «N 


-H   O   <*5   OS   0)   ON    CN 


—  i  ic   ^H  1^.   T-I  OO 


00  O 

01  -* 


01     ^H     01     *-( 


t      g-  &t 

0  Q  Afc  S  <  S  ^^<  <%  0  Z  Q 


RESPIRATION  IX  LITTORINA   1  !<!<(>  RAT  A 


327 


Monthly 
available. 


TABLE  III 

values  of    Littorina  irrorata  at  5°    C  temperature  increments.     ND  indicates  no  data 


Month 

Temperature  "  C 

5-10 

10-15 

15-20 

20-25 

25-30 

,?()-35 

35-40 

40-45 

Oct.  73 

0.9 

2.0 

1.7 

2.5 

1.1 

1.5 

ND 

Dec.  73 

1.6 

0.6 

3.4 

2.0 

4.0 

1.0 

0.9 

ND 

Jan.  74 

1.4 

1.5 

5.3 

3.5 

1.5 

1.5 

1.2 

<0.1 

Feb.  74 

1.8 

1.6 

1.2 

3.0 

1.1 

1.8 

1.9 

<0.1 

Mar.  74 

2.5 

2.5 

2.2 

0.4 

4.7 

1.5 

1.9 

<0.1 

Apr.  74 

3.9 

1.1 

2.1 

1.6 

1.5 

1.5 

1.7 

<0.1 

Mav  74 

1.5             1.7 

2.0 

1.5 

1.7 

2.9 

1.1 

0.6 

Jan.  74 

10.4 

1.9 

2.0 

1.4 

2.7 

1.9 

0.9 

0.1 

Jul.  74                  7.1 

1.2 

2.4 

1.0 

3.8 

2.0 

1.1 

<0.1 

Aug.  74                1.6 

6.8 

1.5 

0.8 

2.6 

2.5 

1.6 

0.7 

Sept.  74 

0.8 

5.1 

0.4 

2.6 

2.9 

2.1 

1.2 

<0.1 

Oct.  74 

2.2 

1.5 

1.5 

2.0 

4.0 

0.7 

1.3 

ND 

Nov.  74 

3.8 

0.2 

41.2 

0.6 

0.6 

2.2 

1.3 

0.4 

Dec.  74 

4.3 

0.5 

11.4 

1.1 

2.3 

1.7 

1.0 

ND 

respiration  rates  occurred  at  night  in  11  of  the  14  months.  Xo  relationships  be- 
tween respiration  rates  and  tidal  cycle  were  evident. 

Comparison  of  respiration  rates  in  study  1  across  all  months  at  the  same  tem- 
perature is  not  possible,  because  of  the  different  experimental  temperatures.  For 
those  months  with  the  same  experimental  temperatures,  as  February  through  April, 
1974,  at  25°  C.  and  May  through  October,  1974,  at  30°  C,  direct  comparisons  can 
be  made.  An  increase  in  oxygen  consumption  rate  occurred  for  both  day  and  night 
readings  for  the  period  February  through  April,  with  the  exception  of  the  day  rate 
in  April.  The  increase  in  oxygen  consumption  rates  continued  for  the  period  May 
through  July,  with  the  exception  of  the  day  rate  in  July.  Oxygen  consumption 
rates  then  declined  for  the  period  August  through  October,  with  the  exception  of 
the  day  rate  in  September.  A  comparison  of  the  rates  during  the  remaining  months 
can  be  made  by  using  the  readings  from  November  and  December  of  1974  and 
January  of  1973,  as  all  have  experimental  temperatures  of  15°  C.  The  decline 
continued  through  November,  increased  in  December  and  remained  at  that  rate  in 
January.  It  may  be  presumed  that  experimental  snails  were  acclimatized  to  field 
temperatures,  making  some  degree  of  seasonal  comparison  possible.  Increased 
respiration  rates,  and  presumably  activity,  increased  with  increasing  temperatures 
from  late  winter  through  midsummer.  Oxygen  consumption  rates  started  declining, 
however,  prior  to  a  corresponding  decline  in  ambient  temperatures.  The  decline  in 
rate  of  consumption  continued  from  late  summer  through  the  fall  months  and  early 
winter  before  starting  to  increase  again. 

In  study  2,  regression  equations  of  Iog1(»  respiration  rate  versus  Iog10  snail  dry 
weight  were  significant  in  only  44  of  121  metabolic  rate-temperature  experiments. 
Monthly  QO2  values,  expressed  as  p.\  O2/(g  dry  wt-hr)  were  therefore  not  norma- 


328 


SHIRLEY,  DENOUX,  AND  STICKLE 


TABLE  IV 

Monthly  Q(>.,  mines  of  Littorina  irrorata  in  vessels  with  sea  water  (controls)  and  in  low  humidity, 
expressed  at  pi  O^/(g  dry  wt-hr)  ±95%  confidence  interval. 


Month 

Temp.  (°  C) 

Control? 

Low  humidity 

—  mucous 

holdfast 

+  mucous  holdfast 

Dec.  73 

15 

257  ±    46 

118  ± 

69 

171  ±209 

Jan.  74 

15 

244  ±    29 

208  ± 

77 

215  db    23 

Feb.  74 

25 

538  ±  101 

677  ± 

160 

Mar.  74 

25 

279  ±    44 

421  ± 

81 

400  ±  758 

Apr.  74 

25 

533  ±  103 

444  ± 

46 

455  ±    73 

May  74 

30 

673  ±    58 

336  ± 

101 

249 

Jim.  74 

30 

795  ±  141 

921  ± 

153 

Jul.  74 

30 

960  ±  227 

841  ± 

167 

Aug.  74 

30 

604  ±  112 

846  ± 

1037 

687  ±  148 

Sep.  74 

30 

865  ±  151 

879  ± 

132 

Oct.  74 

30 

544  ±    81 

498  ± 

48 

368  ±  179 

Nov.  74 

15 

206  ±    35 

300  ± 

29 

Dec.  74 

15 

276  ±    53 

219  ± 

32 

192  ±  155 

lized  from  regression  equations  (Table  II).  The  lowest  temperature  at  which 
rates  were  measured,  5°  C,  had  no  significant  regression  equations.  Similarly,  the 
highest  temperature,  45°  C,  had  only  two  significant  regression  equations.  The 
higher  temperatures,  35°  and  40°  C,  had  the  most  significant  rate  versus  weight 
regression  equations :  9  and  8,  respectively,  for  the  entire  study. 

Although  the  snails  were  inactive  at  5°  C  in  all  months,  with  the  exceptions  of 
May  and  September,  the  rate  of  consumption  was  higher  during  the  colder  months 
and  lower  during  the  warmer  months.  The  trend  was  more  obvious  at  10°  C,  with 
highest  consumption  rates  during  the  colder  months.  The  snails  were  always  in  a 
heat  coma  at  45°  C,  although  no  snails  died  during  the  experiment.  The  Oo  con- 
sumption rates  at  45°  C  frequently  approximated  those  at  5°  C  during  certain 
months.  The  relatively  stable  metabolic  rate  at  intermediate  temperatures  might 
best  be  observed  by  examining  O:0  values  (Table  III).  For  the  months  March 
through  August,  1974  the  Qi0  values  for  the  temperature  range  20-25°  C  are  not 
significantly  different  from  1.0,  indicating  a  temperature  insensitivity.  Values  of 
Qio  at  15-20°  C  and  25-30°  C  average  around  2.0-2.5  over  the  year,  a  normal 
temperature  response  in  poikilotherms. 

Respiration  rates  of  snails  in  vessels  containing  sea  water  and  snails  with  and 
without  mucous  holdfasts  in  low  humidity  for  all  months  are  given  in  Table  IV.  No 
apparent  relationship  between  mucous  holdfast  formation  and  oxygen  consumption 
rates  was  evident.  No  significant  difference  in  respiration  rates  was  found  be- 
tween those  with  and  without  mucous  holdfasts  for  snails  in  low  humidity.  Al- 
though significant  differences  often  occurred  between  respiration  rates  of  snails  in 
low  and  high  humidity,  the  relationship  varied.  During  some  months  the  Q02  values 
of  snails  in  dry  air  would  be  significantly  lower  than  those  of  snails  in  high  humid- 
ity, while  in  other  months  the  inverse  was  true.  In  addition  to  those  nuicous  hold- 


RESPIRATION   IX  LITTORINA  IRRORATA 


329 


TABLE  V 

Energy  budget  for  I.itturina  irroratu/w  the  year  1974.  Dry  weight  index  is  given  in  g/100  g  standard 
animal,  and  caloric,  content  is  modified  from  Bistransin  (1976). 


Month 

Temp.  (°  C) 

Dry  wt 
index 

Mean  dry 
wt  (mg) 

Caloric 
content 
(calories) 

Cal/day 

respired 

Percentage  of 
caloric  content 
respired  per  day 

Jan. 

12 

5.31 

160 

554 

3.6 

0.65 

Feb. 

19 

6.21 

169 

624 

6.5 

1.04 

Mar. 

19 

6.08 

201 

711 

10.2 

1.43 

Apr. 

20 

6.47 

193 

810 

9.3 

1.15 

May 

23 

6.33 

203 

957 

10.9 

1.14 

Jun. 

27 

5.71 

173 

737 

16.1 

2.18 

Jul. 

26 

4.59 

163 

696 

11.2 

1.60 

Aug. 

27 

5.99 

175 

747 

11.7 

1.56 

Sep. 

26 

3.93 

121 

409 

8.4 

2.04 

Oct. 

21 

4.43 

127 

407 

6.7 

1.65 

Nov. 

19 

4.33 

144 

528 

18.3 

3.47 

Dec. 

13 

5.47 

148 

572 

5.2 

0.92 

X 

21 

5.40 

165 

646 

9.8 

1.57 

fasts  whose  formation  were  induced  in  study  3,  the  formation  of  mucous  holdfasts 
also  occurred  in  studies  1  and  2.  A  record  of  these  mucous  holdfasts  was  kept  and 
most  were  formed  during  the  winter  months  at  the  coldest  experimental  tempera- 
tures. None  was  formed  at  temperatures  above  15°  C,  with  the  exception  of  two 
that  were  formed  at  40°  C  in  January,  1974.  Of  74  mucous  holdfasts  that  were 
formed  in  all  determinations  in  studies  1  and  2,  63  were  formed  at  5  or  10°  C,  and 
47  of  those  were  formed  in  January  and  February  of  1974. 

Seasonal  changes  in  the  respiratory  expenditure  of  energy  by  Littorina  irrorata 
are  given  in  Table  V.  The  average  dry  weight  of  the  population  varied  seasonally 
as  the  animals  accumulated  nutrient  reserves  for  spawning,  which  probably  occurred 
in  two  episodes:  June  to  July  and  August  to  September  (Bistransin,  1976).  The 
average  dry  weight  of  snails  cycled  very  closely  with  the  average  dry  weight  index 
of  the  same  population  as  determined  by  Bistransin  (1976).  Caloric  content 
cycled  with  the  dry  weight  indexes  and  average  dry  weight.  The  average  number 
of  calories  respired  per  day  over  the  course  of  the  study  was  9.8,  and  the  percentage 
of  caloric  content  respired  per  day  was  1.57.  Both  values  cycled  seasonally.  The 
calories  respired  per  animal  per  day  cycled  more  closely  with  average  air  tempera- 
ture than  did  the  percentage  of  caloric  content  respired  per  day.  This  was  principally 
due  to  concomitant  changes  in  the  dry  weight  index.  The  high  respiratory  loss  in 
November  is  the  result  of  the  highest  respiration  rate  of  the  year  occurring  at  20 
and  25°  C  during  that  month. 


DISCUSSION 

A  circadian  rhythm  of  oxygen  consumption,  with  higher  rates  of  consumption 
during  the  night,  was  present  in  all  months  of  the  year.  Further  investigation  of 
the  rhythm  of  L.  irrorata  under  various  experimental  conditions  has  demonstrated 


330  SIMKLKY,   DKXOUX,   AXD   STICKLE 

that  light  is  the  phase-setting  factor  and  that  tin-  rhythm  can  he  shifted  according 
to  the  light  regime  (Shirley  and  Findley,  1^78).  It  is  ])r<)])ahle  that  the  rhythm 
reflects  changes  in  the  activity  of  the  snails,  such  as  foraging,  breeding  and  move- 
ments to  more  optimal  conditions.  A  circadian  rhythm  of  feeding-related  activity 
is  not  likely,  as  L.  irrorata  feeds  on  the  exposed  marsh  floor  during  low  tide  (Alex- 
ander, 1976;  Bingham.  1^72a).  A  tidal  rhythm  of  activity  might  therefore  be 
expected;  however,  no  activity  of  L.  irrorata  is  synchronous  with  the  tides  other 
than  its  moving  np  Sparthia  stems  when  covered  by  a  rising  tide  (Bingham,  1972a). 
The  stimulus  to  move  down  the  stems  and  initiate  feeding  is  apparently  increased 
temperature,  but  feeding  will  proceed  only  if  the  marsh  floor  is  exposed  (Bingham, 
1972a).  Moreover,  no  tidal  rhythm  of  respiration  was  detected  in  this  study.  The 
absence  of  tidal  rhythms  of  activity  in  other  supra  and  upper-littoral  littorinid  snails 
has  been  reported  (Zann,  1973).  The  vagaries  of  the  tide  in  the  marshes  of  the 
northern  Gulf  of  Mexico,  with  the  wind  often  having  a  greater  effect  than  lunar 
forces  on  tidal  height,  may  help  explain  the  lack  of  a  tidal  rhythm. 

Respiratory  rhythms  that  have  been  reported  for  other  marine  gastropods  differ 
from  that  of  L.  irrorata.  Sandeen  ct  al.  (1954)  reported  both  diurnal  and  tidal 
rhythms  in  Littorina  littorea  and  Urosulpin.v  cinerens.  In  both  species,  maximal 
respiration  rates  occurred  in  the  hours  following  sunrise  and  sunset.  The  lowest 
rates  occurred  during  the  early  morning  hours,  when  the  highest  rates  were  found 
in  this  study.  Sandison  (1966)  also  reported  a  diurnal  rhythm  of  respiration  by 
L.  sa.ratilis  in  water,  L.  littorea  in  air  and  a  tidal  rhythm  of  respiration  in  the  latter 
species  while  it  was  in  water.  Sandison  (1966)  reported  the  highest  rates  for  L. 
littorea  to  be  between  the  hours  of  800  to  1200.  As  he  only  measured  rates  for  12 
hours  of  the  day,  the  possibility  of  increased  nocturnal  respiration  was  not  examined. 
Both  the  investigations  by  Sandeen  ct  al.  (1954)  and  Sandison  (1966)  measured 
the  rates  of  groups  of  snails  rather  than  the  rates  of  individual  snails,  as  was  done 
in  this  study. 

The  adaptive  significance  of  the  circadian  rhythm  of  respiration  with  respect  to 
the  biology  of  the  snail  is  uncertain  and  warrants  further  investigation.  One  pos- 
sible explanation  is  that  increased  nocturnal  respiration,  and  presumably  activity, 
may  be  related  to  predation.  The  snail  may  be  more  active  at  night  when  it  is  less 
susceptible  to  visually  oriented  predators,  such  as  the  blue  crab,  Callinectes  sapidus 
(Hamilton,  1976).  Certainly  the  amplitude  of  the  rhythm  is  great  enough  to  ob- 
scure relationships  between  oxygen  consumption  and  experimental  variables,  and 
should  be  a  consideration  in  metabolic  experiments. 

A  seasonal  comparison  of  respiration  rates  may  also  be  made  from  the  data  of 
study  1.  The  increase  in  respiration  rates  at  25°  C  from  February  through  April 
and  likewise  from  May  through  July  at  30°  C  may  be  due  to  warm  temperature 
stimulation  of  metabolism.  The  subsequent  decrease  in  respiration  rates  from 
August  through  October  at  30°  C  occurs  prior  to  the  corresponding  decrease  in 
seasonal  temperatures.  The  decrease  in  oxygen  consumption  rates,  indicative  of  a 
seasonal  change  in  metabolism,  is  perhaps  related  to  changes  in  photoperiod 
(Dehnel,  1958).  The  increase  in  respiration  rate  in  the  winter  at  15°  C  prior  to  an 
increase  in  ambient  temperature  is  suggestive  of  cold  temperature  acclimatization. 


RESPIRATION  IN  L1TTORINA  IRRORATA  331 

Tin-  oxygen  consumption  rates  of  L.  irrorata  in  study  2  differ  notably  from 
those  of  temperate  species  (McMahon  and  Russell-l  lunter,  1977;  Sandison,  1967). 
Littorina  irrorata  is  active  near  its  upper  lethal  temperature,  while  the  temperate 
species  are  not.  Also,  the  temperature  at  which  L.  Irrorata  enters  heat  coma  is 
much  higher  than  temperate  littorinids  (Sandison.  1(>(>7;  McMahon  and  Russell- 
Hunter,  1977).  The  two-phased  reaction  of  snails  entering  heat  coma  observed 
by  Sandison  (1967),  which  consisted  of  an  initial  rise  in  respiratory  rate  followed 
by  an  irregular  fall,  was  not  observed  in  L.  irrorata.  Low  O]0  values  for  the  en- 
tire year  are  present  for  the  temperature  range  of  35-40°  C.  Although  activity  of 
the  snails  was  not  inhibited  at  40°  C  during  any  season  of  the  year,  the  low  Q10 
values  suggest  that  40°  C  is  near  the  snails'  upper  limit  of  capacity  adaptation. 
Lewis  (1971)  also  found  that  activity  was  not  inhibited  in  three  species  of  tropical 
intertidal  gastropods  at  37°  C.  A  seasonal  increase  in  the  upper  limit  of  thermal 
tolerance  reported  for  some  temperate  intertidal  molluscs  ( Newell  and  Pye,  1970a) 
was  not  evident  in  L.  irrorata. 

One  of  the  more  interesting  aspects  of  the  respiration  rates  at  various  tempera- 
tures in  study  2  is  a  plateau  in  oxygen  consumption  rates  in  the  temperature  range 
of  20-25°  C,  clearly  demonstrated  by  the  Qi0  values  being  not  significantly  different 
than  1.0  for  the  months  of  March  through  August.  This  is  perhaps  a  thermo- 
neutral  zone,  or  zone  of  temperature  independence,  for  L.  irrorata  during  these 
months.  This  narrow  zone  of  metabolic  homeostasis  occurs  near  the  average  annual 
temperature  at  the  collection  site,  approximately  20°  C.  Likewise,  the  range  of 
the  temperature  independent  zone  approximates  the  average  daily  temperature 
range,  6.5°  C.  Daily  temperature  variations  of  10°  C  or  greater  occurred  less 
than  8%  of  the  time.  Since  oxygen  consumption  rates  were  measured  only  at  dis- 
crete 5°  C  increments,  the  actual  temperature  independent  zone  may  be  several 
degrees  broader  than  the  discernable  5°  C  zone.  The  zone  may  shift  seasonally,  as 
suggested  by  another  group  of  low  OKI  values  present  at  5-10°  and  10-15°  C  in 
October,  1973  through  January,  1974.  Yet  another  set  of  low  Qio  values  is  present 
in  the  15-20°  C  range  in  August  through  October,  1974,  and  in  the  10-15°  C  range 
in  October  through  December,  1974.  The  lack  of  correlation  between  activity  and 
oxygen  consumption  in  this  study  may  have  made  temperature  independent  zones 
less  distinct.  Most  temperature  independent  zones  reported  for  marine  inverte- 
brates have  been  restricted  to  standard  metabolic  rates  (summarized  by  Newell, 
1969,  1973),  although  other  temperature  independent  zones  have  been  reported  for 
routine  metabolic  rates  of  intertidal  snails  (Bertness  and  Schneider,  1976). 

The  ability  of  littorinid  snails  to  attach  themselves  to  a  substrate  by  means  of  a 
mucous  holdfast  and  then  withdraw  into  their  shell  has  been  considered  a  means  by 
which  they  decrease  exposure  to  unfavorable  conditions.  The  relationship  that 
salinity  and  relative  humidity  have  on  mucous  holdfast  formation  in  L.  irrorata  has 
been  investigated  (Bingham,  1972b).  The  effect  of  temperature  on  holdfast 
formation  and  the  effect  that  holdfast  formation  has  on  metabolism  has  not  been 
reported.  Although  it  would  seem  that  inactive  snails  attached  by  a  mucous  hold- 
fast would  have  reduced  metabolic  demands,  no  significant  difference  was  found 
between  the  respiration  rates  of  snails  in  containers  with  sea  water  and  those  with 


332  SHIRLEY,  DENOUX,  AND  STICKLE 

and  without  holdfasts  in  low  humidity.  The  greater  number  of  mucous  holdfasts 
formed  at  low  temperatures  during  the  winter  months  suggests  that  temperature 
must  be  considered  as  an  important  factor  in  inducing  holdfast  formation,  as  well  as 
relative  humidity  and  salinity.  W.  A.  Murphy  (Tulane  University,  personal  com- 
munciation)  has  found  that  snails  form  holdfasts  more  rapidly  at  various  relative 
humidities  at  10°  C  as  compared  to  20  and  30°  C.  The  terrestrial  snail  Otala 
lactca  is  also  more  likely  to  become  dormant  and  form  epiphragms  at  low  relative 
humidities  and  low  temperatures  than  at  high  relative  humidities  and  high  tempera- 
tures (Rokitka  and  Herreid,  1975).  Although  no  metabolic  advantages  were  found 
for  mucous  holdfast  formation  in  this  study,  presumably  holdfast  formation  in  L. 
irrorata  serves  the  same  functions  that  were  reported  by  Vermeij  (1973)  for 
mucous  holdfasts  in  other  littorinid  snails  :  to  reduce  water  loss  and  contact  be- 
tween soft  tissues  and  substrate ;  to  obviate  the  need  for  a  large  water  reservoir ;  and, 
to  increase  the  degree  of  temperature  regulation. 

In  previous  productivity  studies  of  Spartina  marshes,  the  contribution  of  L. 
irrorata  to  community  metabolism  has  been  estimated  (Day  ct  a!.,  1973;  Odum  and 
Smalley,  1959).  Alexander  (1976)  measured  the  egestion  rate  of  L.  irrorata  to  be 
145  g  organic  matter/(m2-yr).  Day,  Smith  and  Gayle  (unpublished  manuscript) 
have  estimated  the  standing  crop  L.  irrorata  in  Louisiana  salt  marshes  to  be  4.9 
g/nr.  If  our  respiration  data,  Alexander's  egestion  data  and  the  Day,  Smith  and 
Gayle's  standing  crop  data  are  used,  and  annual  energy  budget  for  L.  irrorata  is 
calculated  to  be:  182.7  g  organic  matter/(nr'-yr )  total  food  intake,  4.9  g/m2 
standing  crop,  9.8  g/(nr-yr)  net  organic  production,  145  g  organic  matter/ (nr-yr) 
feces  production  and  27.9  g  organic  matter/ (m2-yr)  lost  to  respiration. 


We  wish  to  express  our  thanks  to  Michelle  Bistransin  Ellet,  John  W.  Day,  Jr., 
and  William  A.  Murphy  for  permitting  us  to  use  their  unpublished  data.  WTe  also 
thank  Thomas  H.  Dietz  for  reviewing  the  manuscript  and  Alison  Hanson,  Deborah 
French,  David  Randall,  Karen  Westphal.  and  Jan  Judice  for  technical  assistance. 
The  investigation  was  supported  in  part  by  the  Petroleum  Refiners  Environmental 
Council  of  Louisiana. 

SUMMARY 

1.  Respiration  rates  of  Littorina  irrorata  were  measured  monthly  for  the  period 
from  October,  1973,  through  December,  1974.     The  study  consisted  of  three  main 
parts :   first,  hourly  measurements  of  respiration  rates  at  ambient  field  air  tempera- 
ture over  a  36  hr  time  period ;  secondly,  one-hour  measurements  of  respiration  rates 
at  5°   C  temperature  increments  from   5°   to  45°    C  during  daylight  hours;  and 
thirdly,  one-hour  measurements  of  respiratory  rates  under  conditions  conducive  to 
mucous  holdfast  formation.     Respiration  rates  were  measured  with  a  Gilson  res- 
pi  rometer  using  standard  manometric  techniques. 

2.  A  diurnal  rhythm  of  respiration  was  found  for  1 1  of  the  14  months.    Respira- 
tion rates  during  the  night  were  significantly  higher  than  during  the  day. 


RESPIRATION  IN  LITTORINA  IRRORATA  333 

3.  Snails  were  in  thermal  stress  at  5°  C  and  45°  C  and  their  respiration  rates 
were  depressed.    Respiration  rates  at  10°  C  were  highest  during  the  colder  months, 
demonstrating  inverse  cold  temperature  acclimatization.     The  Oio  for  the  tempera- 
ture range  20-25°    C  were  not   significantly   different   from    1.0  for   the   months 
March   through   August,    suggesting   thermal   insensitivity   or   the   presence   of   a 
thermo-neutral  zone. 

4.  No  apparent  relationship  between  mucous  holdfast  formation  and  oxygen  con- 
sumption was  evident.  Mucous  holdfasts  were  formed  most  frequently  during  the 
winter  months  at  the  coldest  experimental  temperatures. 

5.  An  annual  energy  budget  of  L.  irrorata  is  calculated. 

LITERATURE  CITED 

ALDRICH,  J.  C.,  1975.  On  the  oxygen  consumption  of  the  crabs  Cancer  pagurus  (L.)  and 
Mala  sqmnado  (Herbst.).  Coinp.  Biochcm.  Physiol.,  50A  :  223-228. 

ALEXANDER,  S.  K.,  1976.  Relationship  of  macrophyte  detritus  to  the  salt  marsh  periwinkle, 
Littorina  irrorata  Say.  Ph.D.  dissertation,  Louisiana  State  University,  Baton  Rouge, 
114  pp.  (Diss.  Abstr.,  37:  5506-B  ;  order  no.  77-10.353.) 

BARNES,  H.,  AND  M.  BARNES,  1969.  Seasonal  changes  in  the  acutely  determined  oxygen  con- 
sumption and  effect  of  temperature  for  three  common  cirripedes,  Balanus  balanoidcs 
(L.),  B.  balanus  (L.)  and  Chthamahts  stellatus  (Poli).  /.  Exp.  Mar.  Biol.  Ecol,  4: 
36-50. 

BEQUAERT,  J.  C.,  1943.    The  genus  Littorina  in  the  western  Atlantic.     Johnsonia,  1(7):  1-27. 

BERTNESS,  M.  D.,  AND  D.  E.  SCHNEIDER,  1976.  Temperature  relations  of  Puget  Sound  Thaids 
in  reference  to  their  intertidal  distribution.  Veligcr,  19 :  47-58. 

BINGHAM,  F.  O.,  1972a.  The  influence  of  environmental  stimuli  on  the  direction  of  movement 
of  the  supralittoral  gastropod  Littorina  irrorata.  Bull.  Mar.  Sci.,  22:  309-335. 

BINGHAM,  F.  O.,  1972b.  The  mucous  holdfast  of  Littorina  irrorata  and  its  relationship  to  rela- 
tive humidity  and  salinity.  Veligcr,  15  :  48-50. 

BISTRANSIN,  M.  E.,  1976.  The  reproductive  physiology  of  the  prosobranch  snail  Littorina  ir- 
rorata (Say,  1822).  Master's  thesis,  Louisiana  State  University,  Baton  Rouge, 
Louisiana,  53  pp. 

COLEMAN,  N.,  1976.  Aerial  respiration  of  nerites  from  the  Northeast  coast  of  Australia.  Aust. 
J.  Mar.  Frcslra'atcr  Res.,  27  :  455-466. 

CRISP,  D.  J.,  1971.  Energy  flow  measurements.  Pages  197-279  in  N.  A.  Holmes  and  A.  D. 
Mclntyre,  Eds,  IBP  Handbook  No.  16:  Methods  for  the  Study  of  Marine  Benthos. 
Blackwell  Scientific  Publications,  Oxford. 

DAY,  J.  W,  JR.,  G.  SMITH,  P.  R.  WAGNER,  AND  W.  C.  STOWE,  1973.  Community  structure  and 
carbon  budget  of  a  salt  marsh  and  shallow  bay  cstuarine  system  in  Louisiana.  Louisiana 
State  University  Sea  Grant  Publication,  72-04,  Baton  Rouge,  80  pp. 

DEHNEL,  P.  A.,  1958.  Effect  of  photoperiod  on  the  oxygen  consumption  of  two  species  of 
intertidal  crabs.  Nature,  181:  1415-1417. 

HAMILTON,  P.  V.,  1976.  Predation  on  Littorina  irrorata  (Mollusca:  Gastropoda)  by  CaUincctes 
sapidus  (Crustacea:  Portunidae).  Bull.  Mar.  Sci.,  26:  403-409. 

HUEBNER,  J.  D.,  1973.  The  effect  of  body  size  and  temperature  on  the  respiration  of  Poliniccs 
duplicatus.  Coinp.  Biochcm.  Physiol.,  44A  :  1185-1197. 

LEWIS,  J.  B,  1971.  Comparative  respiration  of  some  tropical  intertidal  gastropods.  /.  Exp. 
Mar.  Biol.  Ecol. ,6:  101-108. 

MCFARLAND,  W.  N.,  AND  P.  E.  PiCKENS,  1965.  The  effects  of  season,  temperature  and  salinity 
on  standard  and  active  oxygen  consumption  of  the  grass  shrimp  Palaemonetcs  vulgaris 
(Say).  Can.  J.  ZooL,  43  :  571-585. 

McLusKY,  D.  S,  1973.  The  effect  of  temperature  on  the  oxygen  consumption  and  nitration 
rate  of  Chlamys  (Acquipectcn}  opcrcularis  (L.)  (Bivalvia).  Ophelia,  10:  141-154. 


334  SHIRLEY,  DENOUX,  AND  STICKLE 

McMAHON,    R.    F.,    AND   W.    D.    RUSSELL-HUNTER,    1977.     Temperature    relations    of    aerial 

and  aquatic  respiration  in  six  littoral  snails  in  relation  to  their  vertical  zonation.    Biol. 

Bull.,  152 :  182-198. 
NEWELL,   R.   C.,   1969.     Effect  of  fluctuations   in   temperature  on  the  metabolism   of  intertidal 

invertebrates.   Am.  Zoo/.,  9:  293-307. 
NEWELL,  R.  C.,  1973.     Factors  affecting  the  respiration  of  intertidal  invertebrates.    Am.  ZooL, 

13 :  513-528. 
NEWELL,    R.    C.,    1976.     Adaptations   to    intertidal    life.      Pages    1-82    in    R.    C.    Newell,    Ed., 

Adaptation  to  environment.    Butterworths,  London. 
NEWELL,  R.  C.,  AND  H.  R.  NORTHCROFT,  1965.     The  relationship  between  cirral  activity  and 

oxygen  uptake  in  Balamis  balanoidcs.    J.  Mar.  Biol.  Assoc.  U.  K.,  45 :  387-403. 
NEWELL,  R.  C.,  AND  V.  I.  PYE,  1970a.     Seasonal  changes  in  the  effect  of  temperature  on  the 

oxygen  consumption  of  the  winkle  Littorina  littorea  (L.)  and  the  mussell  Mytilus  edulis 

L.  Comp.  Biochcm.  PhysioL,  34:  367-383. 
NEWELL,  R.  C,  AND  V.  I.  PYE,  1970b.     The  influence  of  thermal  acclimation  on  the  relation 

between  oxygen  consumption  and  temperature  in  Littorina  irrorata   (L.)   and  Mytilus 

edulis  L.  Comp.  Biochem.  PhysioL,  45 :  385-397. 
NEWELL,  R.  C.,  AND  V.  I.  PYE,  1971.     Quantitative  aspects  of  the  relationship  between  oxygen 

consumption,  body  size  and  summated  tissue  metabolism  in  the  winkle,  Littorina  littorea. 

J.  Mar.  Biol.  Assoc.  U.K.,  51:  315-338. 

NEWELL,  R.  C,  AND  A.  ROY,  1973.     A  statistical  model  relating  the  oxygen  consumption  of  a 
mollusc   (Littorina  littorea}  to  activity,  body  size  and  environmental  conditions.  PhysioL 

ZooL,  46 :  253-275. 
NEWELL,  R.  C.,  W.  WIESER,  AND  V.  I.  PYE,  1974.     Factors  affecting  oxygen  consumption  in 

the  woodlouse  Porcellio  scaber  Latr.  Oecologia,  16:  31-51. 
PYE,  V.  I.,  AND  R.  W.  NEWELL,  1973.     Factors  affecting  thermal  compensation  in  the  oxidative 

metabolism  of  the  winkle  Littorina  littorea.    Neth.  J.  Sea.  Res.,  7:  411-419. 
ODUM,  H.  T.,  AND  A.  E.  SMALLEY,  1959.     Comparison  of  population  energy  flow  of  a  herbivor- 
ous and  a  deposit-feeding  invertebrate  in  a   salt  marsh  ecosystem.    Proc.  Nat.  Acad. 

Sci.  US.A.,4S:  617-622. 
ROKITKA,   M.  A.,  AND   C.  F.   HERREID  ii,   1975.     Formation  of  epiphragms   by  the   land   snail 

Otala  lactea  (Muller)  under  various  environmental  conditions.   Nautilus,  89 :  27-32. 
RUSSELL-HUNTER,  W.,   1964.     Physiological  aspects  of  ecology  in  nonmarine  molluscs.    Pages 

83-126  in  K.   M.  Wilbur  and  C.   M.  Yonge,   Eds.,   Physiology   of  Mollusca,   Vol.   I. 

Academic  Press,  New  York  and  London. 
SANDEEN,  M.  L,  G.  C.   STEPHENS,  AND  F.  A.  BROWN,  JR.,   1954.     Persistent  daily  and  tidal 

rhythms  of  oxygen  consumption  in  two  species  of  marine  snails.    PhysioL  ZooL,  27 : 

350-356. 
SANDISON,  E.  E.,  1966.     The  oxygen  consumption  of  some  intertidal  gastropods  in  relation  to 

zonation.  /.  ZooL,  149:  163-173. 
SANDISON,  E.  E.  1967.     Respiratory  response  to  temperature  and  temperature  tolerance  of  some 

intertidal  gastropods.    /.  E.vp.  Mar.  Biol.  EcoL,  1 :  271-281. 
SHIRLEY,  T.  C.,  AND  A.  M.  FINDLEY,  1978.     Circadian  rhythm  of  oxygen  consumption  in  the 

marsh  periwinkle,  Littorina  irrorata  (Say,  1822).     Comp.  Biochem.  PhysioL,  in  press. 
SNEDECOR,  G.  W.,  AND  W.  G.  COCHRAX,  1971.     Statistical  methods,  6th  Ed.    The  Iowa  State 

University  Press,  Ames,  593  pp. 
SUBRAHMANYAM,  C.  B.,  W.  L.  KRUCZYNSKi,  AND  S.  H.  DRAKE,  1976.       Studies  on  the  animal 

communities  in  two  north   Florida   salt  marshes.    Part   II.     Macroinvertebrate   com- 
munities. Bull.  Mar.  Sci.,  26:  172-195. 
VERMEIJ,  G.  J.,  1973.    Morphological  patterns  in  high  intertidal  gastropods :  adaptive  strategies 

and  their  limitations.  Mar.  Biol.,  20:  319-346. 

ZANN,  L.  P.,  1973.     Relationships  between  intertidal  zonation  and  circatidal  rhythmicity  in  lit- 
toral gastropods.  Mar.  Biol.,  18  :  243-250. 


Reference:  Biol  Bull.,  154 :  335-347.    (April,  1978) 


TRANSEPIDERMAL  ACCUMULATION  OF  NATURALLY  OCCURRING 

AMINO  ACIDS  IN  THE  SAND  DOLLAR, 

DENDRASTER  EXCENTRICUS  x 

GROVER  C.  STEPHENS,  MARVA  J.  VOLK,  STEPHEN  H.  WRIGHT 

AND  PETER  S.  BACKLUND 

Department  of  Developmental  and  Cell  Biology,  University  of  California,  Irvine, 

Irvine,  California  92717 

Echinoids  and  asteroids  exhibit  a  full  range  of  feeding  habits  including  car- 
nivores, herbivores,  detritus  feeders  and  filter  feeders.  However,  distribution  of 
nutrients  derived  from  digestion,  whatever  the  feeding  habit,  appears  to  be  slow 
and  incomplete.  Ferguson  (1970)  injected  a  mixture  of  14C-labeled  amino  acids 
into  the  stomach  and  perivisceral  coelom  of  the  starfish,  Echinaster,  and  followed 
the  subsequent  distribution  of  labelled  material  using  autoradiography.  Transloca- 
tion  of  nutrients  throughout  visceral  and  subepidermal  regions  was  observed,  but 
no  labelling  of  epidermal  tissue  was  evident  at  the  end  of  75  days.  Conversely 
(Ferguson,  1967),  "CTabeled  amino  acids  supplied  in  the  ambient  medium  were 
incorporated  into  epidermal  tissues  of  starfishes,  but  there  was  little  or  no  export 
of  label  from  epidermis  to  subepidermal  or  visceral  tissues.  A  comparable  barrier 
to  distribution  of  nutrients  between  epidermal  and  visceral  tissues  is  discussed  in 
the  work  Pequignat  (1969,  1970),  Pequignat  and  Pujol  (1968),  and  Pearse  and 
Pearse  (1973)  employing  various  echinoids  and  asteroids.  The  barrier  is  not 
necessarily  complete.  Slow  translocation  across  the  barrier  is  reported  by  some 
investigators.  These  observations  agree  well  with  the  morphology  of  these  echino- 
derm  groups ;  the  "circulatory"  systems,  though  complex,  do  not  appear  to  provide 
a  well-organized  morphological  substrate  for  distribution  of  material  to  the  epi- 
dermis. 

Stephens  and  Schinske  (1961)  showed  net  influx  of  glycine  from  a  rather  con- 
centrated solution  into  two  species  of  starfishes.  Since  that  time,  a  number  of  in- 
vestigators have  studied  uptake  of  amino  acids  in  echinoderms  (e.g.,  Stephens  and 
Virkar,  1966;  Fontaine  and  Chia,  1968;  Clark,  1969;  Dixit,  1973;  Ahearn  and 
Townsley,  1975).  Most  of  this  work  used  "CTabeled  substrates  and  described 
kinetics  of  influx  by  radiochemical  techniques  and/or  distribution  of  labeled  material 
by  autoradiography.  However,  Ferguson  (1971)  showed  by  direct  chemical 
determination  that  there  was  a  net  influx  of  amino  acids  from  a  medium  concentra- 
tion of  37.5  /AM  into  ten  different  species  of  starfishes  from  the  Puget  Sound  area. 
This  work  did  not  permit  estimation  of  rates  of  net  influx  but  established  the 
capacity  of  several  of  the  forms  employed  to  reduce  ambient  amino  acid  concentra- 
tions to  extremely  low  levels  at  the  end  of  a  six-hour  incubation  period. 

There  is,  thus,  considerable  evidence  for  uptake  and  utilization  of  amino  acids  by 
epidermal  tissues  in  echinoids  and  asteroids.  There  is  also  considerable  evidence 

1  Supported  by  Grant  No.  OCE  76-12183  from  the  National  Science  Foundation. 

335 


336  STEPHENS,  VOLK,  WRIGHT,  AND  BACKLUND 

that  distribution  of  nutrients  from  the  digestive  system  to  superficial  tissues  is  very 
slow.  This  has  led  Ferguson  (1970)  to  suggest  that  epidermal  tissues  may  derive 
much  of  their  sustenance  by  direct  influx  of  nutrients  from  the  environment  inde- 
pendent of  the  nutrition  of  visceral  and  subepidermal  tissues.  Pequignat  (1970) 
frames  a  similar  hypothesis,  adding  the  possibility  of  cutaneous  digestion  and  as- 
similation of  larger  organic  substrates  to  account  for  support  of  epidermal  struc- 
tures. 

The  sand  dollar,  D  end-raster  excentrlcus,  was  selected  as  an  experimental  organ- 
ism for  two  reasons.  First,  there  are  large  populations  of  this  organism  readily 
available  for  study  and  the  feeding  behavior  has  been  well  described  (Timko,  1976). 
Secondly,  we  wanted  to  work  with  an  organism  which  lives  in  or  on  a  soft  substrate, 
since  the  occurrence  and  distribution  of  amino  acids  in  such  habitats  has  already 
been  analyzed  (Stephens,  1975)  in  relation  to  the  nutrition  of  annelid  infauna. 

In  the  present  work,  data  are  presented  on  the  following :  kinetics  of  influx  of 
14C-labeled  amino  acids  from  solution,  kinetics  of  net  influx  of  amino  acids,  levels 
of  amino  acid  present  in  the  microenvironment,  availability  of  naturally  occurring 
amino  acids  as  assessed  by  net  influx,  estimates  of  energy  metabolism  of  whole 
animals,  and  estimates  of  energy  metabolism  in  isolated  portions  of  the  test.  Col- 
lectively, this  information  allows  an  estimate  of  possible  contributions  of  trans- 
epidermal  transport  of  amino  acids  to  the  support  of  epidermal  tissues. 

MATERIALS  AND  METHODS 

Animals  and  sediment  samples  were  obtained  from  two  locations.  Most  of 
the  material  studied  was  obtained  from  a  shallow  water  population  in  a  lagoon  at 
Point  Mugu  Naval  Base  near  Port  Hueneme,  California.  Some  samples  were  taken 
from  a  population  off  Newport  Beach  at  a  depth  of  approximately  5  meters.  Ani- 
mals were  maintained  in  aerated  sea  water  at  a  temperature  of  15°  C.  Sediment 
cores  were  taken  using  plexiglass  coring  tubes  (25  mm  internal  diameter)  and  were 
analyzed  promptly.  Interstitial  water  from  sediment  was  expressed  through  a 
Millipore  filter  (0.45  /xin)  under  nitrogen  at  10-20  psi. 

Influx  rates  of  amino  acids  were  determined  by  measuring  the  disappearance  of 
radioactivity  from  a  solution  containing  the  14C-labeled  compound.  Solutions  were 
prepared  in  artificial  sea  water  (Cavanaugh,  1956)  prepared  from  reagent  grade 
salts.  Radioactivity  was  initially  20  ^Ci/liter  (2-5  X  10~T  moles/liter  depending  on 
specific  activity)  with  12C-amino  acid  added  to  obtain  the  desired  concentration. 
Radioactivity  was  measured  using  a  scintillation  counter;  samples  of  0.5  ml  were 
added  to  a  toluene-based  cocktail  containing  a  detergent  to  solubilize  the  sample. 
Samples  were  acidified  to  drive  off  COo.  Volumes  to  which  animals  were  exposed 
ranged  from  50  to  200  ml ;  air  was  bubbled  slowly  through  the  vessel  to  provide 
aeration  and  circulation. 

Net  flux  of  amino  acids  was  followed  using  fluorescamine  (North,  1975)  to 
determine  primary  amines  in  solutions  to  which  the  animals  were  exposed.  After 
sample  preparation,  fluorescence  was  measured  using  a  Perkin-Elmer  spectro- 
photofluorometer  with  an  excitation  wavelength  of  390  nm  and  an  emission  wave- 
length of  480  nm.  This  procedure  was  also  used  for  estimating  levels  of  naturally 
occurring  primary  amines  in  interstitial  water  of  sediment  samples.  Some  of  the 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER 


337 


20  r 


prmary    amne 


—  o 


O-5  l-O  1-5 

TIME  (hrs) 


2-O 


2-5 


3-0 


FIGURE  1.  Removal  of  serine  from  100  ml  of  a  20  /XM  solution  by  a  specimen  of  Dcndrastcr. 
The  solid  line  shows  decrease  in  radioactivity  with  time;  the  broken  line  shows  decrease  in 
primary  amines  estimated  by  the  fluorescamine  procedure. 

latter  measurements  were  repeated  using  an  independent  procedure  based  on  o- 
phthalaldehyde  (OPA).  The  procedure  is  based  on  that  of  Alendez  and  Gavilanes 
(1976)  but  employs  a  lower  concentration  of  OPA  (0.03  mg/ml  rather  than  0.3 
mg/ml).  It  proved  to  be  necessary  to  read  samples  and  standards  at  a  constant 
time  after  preparation. 

Amino  acids  present  in  interstitial  water  before  and  after  exposure  to  the  ani- 
mals were  identified  by  thin  layer  chromatography.  Seawater  samples  (5-10  ml) 
were  desalted  on  Dowex-50,  eluted  off  the  column  with  3  N  NH4OH  and  chromato- 
graphed  as  described  by  Clark  (1968)  with  the  following  modifications.  Spots 
(1-5  /xl)  were  applied  to  the  chromatogram  using  a  glass  capillary  drawn  to  a  tip 
diameter  of  approximately  50  /xm  which  minimized  spot  diameter.  Chromatograms, 
10  cm  X  10  cm,  rather  than  20  cm  X  20  cm  as  supplied  (Polygram  CEL  300), 
were  developed  for  approximately  one  hour  in  each  dimension  without  scoring 
following  the  first  solvent  system. 

Spots  were  located  using  OPA  as  a  location  reagent  as  follows.  An  OPA 
stock  solution  is  prepared  several  hours  prior  to  use  by  dissolving  30  mg  OPA  in 
200  ml  glass  distilled  water  and  adding  a  drop  of  1  N  NaOH.  This  stock  is  stable 
for  one  to  two  days.  Just  prior  to  use,  the  spray  mixture  is  prepared  consisting  of 
20  ml  OPA  stock,  20  ml  absolute  ethanol,  0.2  ml  triethylamine  and  0.01  ml  2- 


338 


STEPHENS,  VOLK,  WRIGHT,  AND  BACKLUND 


TABLE  I 


Rates  of  disappearance  of  amino  acids  expressed  as  nmoles /  (hr  •  cmz) . 
both  surfaces  are  combined.     Average  diameter  4.1-7.6  cm. 


Data  for  aboral  surface  and  for 


Mean  rate 

Standard  deviation 

N 

ala 

42.0 

2.7 

2 

asp 

16.8 

2.3 

3 

gly 

45.8 

5.0 

6 

glu 

6.1 

1.0 

5 

lys 

25.4 

3.8 

5 

ser 

49.5 

5.5 

8 

val 

63.0 

4.6 

3 

naturally  occurring 

primary  amines 

24.9 

7.6 

5 

mercaptoethanol.  The  spray  mixture  is  stable  for  several  hours.  It  is  applied 
using  a  mist  sprayer;  approximately  5  ml  suffices  for  a  10  cm  X  10  cm  plate.  One 
to  ten  minutes  after  spraying,  chromatograms  are  examined  under  a  long  wavelength 
UV-lamp.  Most  spots  intensify  on  drying,  but  lysine  fades.  Detection  levels  for 
most  amino  acids  range  from  20  to  50  picomoles.  However,  200  to  300  picomoles 
are  required  for  location  of  some  hydrophobia  amino  acids  (e.g.,  val,  leu,  ile). 
Chromatograms  were  photographed  through  a  yellow  filter  (455  nm)  using  Kodak 
Tri-X  film  and  diafine  development  (ASA  1600)  at  f5.6  and  0.5  second  exposure. 
Schiltz,  Schnackerz  and  Gray  (1977)  have  recently  described  a  comparable  pro- 
cedure. When  chromatograms  are  prepared  as  described,  ninhydrin  can  also  be 
used  as  a  location  reagent  with  detection  levels  in  the  range  of  a  few  hundred  pico- 
moles. 

Oxygen  consumption  was  measured  using  a  YSI  oxygen  electrode.  Measure- 
ments of  oxygen  consumption  as  well  as  studies  of  influx  and  net  flux  of  amino 
acids  were  carried  out  at  a  temperature  of  20°  C. 


RESULTS 

Figure  1  presents  the  results  of  a  typical  set  of  observations.  A  sand  dollar,  7.2 
cm  average  diameter,  was  placed  in  100  ml  of  artificial  sea  water  to  which  serine 
had  been  added  at  a  concentration  of  20  /^moles/liter.  The  solution  also  contained 
2  /xCi  of  14C-serine  (UL).  As  indicated  in  the  figure,  radioactivity  decreased 
rapidly  with  time  as  did  the  total  primary  amine  in  the  solution  estimated  by  the 
fluorescamine  reaction.  Disappearance  followed  first  order  exponential  kinetics  for 
the  first  hour,  and  the  two  curves  are  virtually  identical.  The  rate  of  entry  (dis- 
appearance of  primary  amine)  at  the  initial  concentration  of  20  JU.M  was  2.15 
/mioles/hr.  It  proved  to  be  best  to  relate  rates  to  the  surface  area  of  the  animals 
and  express  them  as  nmoles/ (hr- cm2).  This  gave  consistent  results  over  the 
considerable  size  range  of  animals  examined.  For  the  case  presented  in  Figure  1, 
uptake  of  serine  from  a  20  /*M  solution  proceeded  at  a  rate  of  53  nmoles/ (hr- cm2). 
If  influx  of  14C  is  expressed  in  the  same  units,  the  rate  from  a  20  /J.M  solution  is 
56.7  nmoles/ (hr- cm2). 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER 


339 


3  5 

O 
>=  4 


u 
O 


primary   amme 
'  C-glydne 


O-5        l-O 


1-5       2-O      2-5 
TSME  (hrs) 


3-0 


25-0 


FIGURE  2.  Removal  of  glycine  from  100  ml  of  a  5  /UM  solution  by  two  specimens  of  Den- 
draster.  Data  points  for  the  two  individuals  are  solid  and  open  circles  and  solid  and  open 
squares  respectively. 


Results  of  experiments  of  this  kind  were  quite  repeatable.  When  animals  were 
supported  on  glass  rods  to  facilitate  circulation  of  solution  across  the  oral  surface, 
observed  rates  of  disappearance  of  an  added  amino  acid  were  approximately  doubled. 
Thus  both  surfaces  of  the  animals  seem  to  participate  equally  in  removal  of  amino 
acid  from  dilute  solution.  Table  I  presents  rates  of  disappearance  of  various  amino 
acids.  All  rates  are  expressed  as  nmoles/(hr-cm2)  at  an  ambient  concentration  of 
20  fj.M.  It  should  be  noted  that  the  determinations  of  14C  and  of  primary  amines  di- 
verge at  low  concentrations  (Fig.  1).  Figure  2  shows  data  for  two  different  ani- 
mals offered  14C-glycine  at  5  /^moles/liter.  Fluorescent  material  declined  over  the 
course  of  two  hours  to  approximately  1  /^mole/liter  (glycine-equivalent  concentra- 
tion) and  then  slowly  increased  to  levels  of  5-7  p.M  over  the  course  of  the  ensuing 
23  hours.  If  animals  were  placed  in  a  small  volume  of  artificial  sea  water  with  no 
added  amino  acid,  primary  amines  slowly  increased  and  stabilized  at  similar  final 
concentrations.  These  same  levels  were  found  in  the  aerated  water  in  which  groups 
of  animals  were  kept  over  a  period  of  days. 

Figure  2  also  shows  that  small  amounts  of  radioactivity  (6-8%)  remained  in 
the  solution  25  hours  after  14C-glycine  was  supplied.  In  the  case  of  other  amino 
acids,  for  example  serine,  this  effect  was  quite  pronounced  with  as  much  as  15-20% 
of  initial  radioactivity  persisting  in  solution  at  the  end  of  24  hours.  The  radio- 
activity does  not  appear  to  be  in  the  form  of  serine ;  about  60%  of  the  activity 
passes  through  a  Dowex-50  column  in  the  acid  form,  and  TLC  shows  several  spots 
that  are  unidentified  but  do  not  react  as  primary  amines. 


340 


STEPHENS,  VOLK,  WRIGHT,  AND  BACKLUND 


w 

"o 


1-2 


l-O 


0-8 


0-6 


O4 


O-2 


2OO 


10 


50 


2OO 


FIGURE  3.  Removal  of  glycine  from  solution  by  Dendraster  as  a  function  of  ambient  con- 
centrations. The  insert  graph  presents  rates  determined  at  the  concentrations  indicated ;  the 
curve  is  a  hyperbola  fitted  to  the  kinetic  constants.  The  larger  graph  is  a  Woolf  plot  of 
the  data.  Kt  is  74  /*M  Fmax  is  215  nmoles/ (hr •  cm2) . 


Figure  3  presents  data  relating  influx  (measured  by  disappearance  of  14C)  to 
glycine  concentration.  The  insert  is  a  plot  of  influx  as  a  function  of  ambient  con- 
centration. Kinetic  constants  were  evaluated  from  the  Woolf  plot  presented  in 
Figure  3.  The  Kt  was  74  /AM  and  the  Vma.\  was  215  nmoles/ (hr- cm2). 

Dendraster  does  not  take  up  glycylglycine  from  dilute  solution.  Animals  were 
incubated  for  24  hours  in  20  /AM  glycylglycine  with  a  trace  amount  of  14C-glycine. 
Radioactivity  in  the  medium  decreased  rapidly,  as  expected.  Fluorescamine  posi- 
tive material  (expressed  as  equivalent  glycylglycine  concentration)  increased 
slightly  during  early  incubation,  presumably  reflecting  efflux  of  unknown  primary 
amines.  After  24  hours,  levels  had  decreased  to  about  75%  of  the  original  con- 
centration. This  may  reflect  microbial  activity  or  very  slow  uptake.  In  any  case, 
uptake  of  glycylglycine  either  does  not  occur  at  all  or  is  so  slow  as  to  be  insignificant 
compared  to  uptake  of  neutral  amino  acids. 

Naturally  occurring  primary  amines  in  interstitial  water  were  determined  using 
the  fluorescamine  technique  and,  in  some  cases,  OPA.  Determinations  using  the 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER  341 

two  procedures  were  in  good  agreement.  Sediment  cores  were  divided  into  3  cm 
zones  from  the  surface  downward ;  water  was  expressed  through  a  Millipore  filter 
under  NT2,  and  concentrations  of  primary  amines  expressed  as  glycine-equivalent 
concentration.  \Ye  do  not  believe  that  the  cores  taken  at  5  meters  depth  were  un- 
disturbed. They  showed  interstitial  concentrations  of  17  and  23  /mioles  amines/ 
liter,  respectively,  in  the  top  3  cm.  Core  samples  could  be  taken  in  the  immediate 
vicinity  of  the  shallow  water  population  with  minimum  disturbance  of  sediment 
organization.  The  samples  showed  great  variability  in  primary  amine  content. 
Fifteen  samples  gave  an  averaged  value  of  115  /AM  in  the  interstitial  water 
of  the  top  3  cm  of  the  cores  with  a  standard  deviation  of  60  /AM.  The  range 
was  17-244  /AM.  Stephens  (1975;  also  reports  considerable  variability  in  primary 
amine  concentration  in  sediment  cores.  In  general,  primary  amine  concentration 
decreased  with  depth,  also  in  agreement  with  Stephens  (1975)  and  Crowe,  Dickson, 
Otto,  Colon  and  Farley  (1977),  though  there  were  two  cores  which  showed  an 
increase  at  the  3-6  cm  and  6-9  cm  zones. 

Observations  were  carried  out  on  rates  of  influx  using  samples  of  naturally  oc- 
curring primary  amines  from  both  collection  sites.  Although  the  samples  were  ex- 
pressed from  sediment  which  was  collected  as  carefully  as  possible,  it  is  likely  that 
they  were  somewhat  diluted  during  collection.  Also,  a  period  of  several  hours 
elapsed  before  it  was  possible  to  obtain  interstitial  water  from  sediment  collected 
at  the  shallow  water  site.  The  initial  concentrations  for  the  two  sets  of  observations 
were  14  /AM  (for  the  deeper  population)  and  33  /AM.  The  results  are  presented  in 
Table  I,  recalculated  to  present  rates  from  an  ambient  concentration  of  20  /AM  to 
facilitate  comparison  with  rates  for  known  amino  acids.  The  correction  was  made 
assuming  a  linear  relation  between  ambient  concentration  and  influx  over  the 
relevant  range  (14—33  /x,M). 

Figure  4  presents  photographs  of  TLC,  including  a  standard  and  samples  of  sea 
water  before  and  after  exposure  to  a  sand  dollar  for  24  hours.  The  standard  con- 
tained 250  picomoles  of  each  amino  acid.  The  sea  water  to  which  the  animal  was 
exposed  was  interstitial  water  which  initially  contained  33  /AM  primary  amine  as 
estimated  by  the  fluorescamine  procedure.  Final  concentration  was  7  /AM.  Desalted 
samples  representing  125  /A!  of  the  interstitial  water  before  and  after  exposure  were 
spotted  and  chromatographed.  Figure  4  illustrates  the  marked  decrease  in  neutral 
amino  acids  at  the  end  of  the  exposure  period.  A  larger  amount  of  the  post-exposure 
sample  was  spotted  and  chromatographed ;  spots  were  more  intense,  but  the  pattern 
was  the  same  as  that  illustrated.  Neither  the  initial  nor  the  final  sample  chromato- 
graph  in  Figure  4  should  be  interpreted  as  a  complete  inventory  of  primary  amines 
in  the  interstitial  water.  Only  70-85%  of  primary  amine  as  estimated  by  the 
fluorescamine  procedure  is  retained  on  passage  through  a  Dowex-50  column  in 
the  acid  phase  and  subsequently  eluted  with  XH4OH.  whereas  the  retention  of  a 
standard  mixture  of  amino  acids  in  sea  water  is  virtually  complete.  Thus,  some  of 
the  naturally  occurring  primary  amine  is  not  behaving  as  do  most  amino  acids,  is 
not  present  in  our  desalted  sample,  and  hence  is  not  represented  on  the  TLC.  As 
an  example,  taurine  reacts  with  fluorescamine  but  is  not  retained  on  a  Dowex 
column.  However,  primary  amines  which  pass  through  the  column  were  not 
identified. 


342 


STEPHENS,  VOLK,  WRIGHT,  AND  BACKLUND 


80 

6Q         70 

6o  7° 

6n  7° 

30  4o  5o 

3o  4o  5o 

u 
405o 

°02 

2° 

"0 

; 
1° 

L 

100     90 

10o 

FIGURE  4.  Thin  layer  chromatograms  developed  with  OPA  (see  text  for  procedure). 
Amino  acids  are  coded  as  (1)  arg,  (2)  lys,  (3)  asp,  (4)  gly,  (5)  ser,  (6)  glu,  (7)  ala,  (8) 
val,  (9)  his,  (10)  orn,  (11)  gin.  A  is  a  standard  containing  250  picomoles  of  each  of  the  first 
8  amino  acids,  made  in  artificial  sea  water,  desalted  and  run.  B  is  a  sample  of  interstitial  water ; 
total  primary  amine  content  is  approximately  4.1  nanomoles.  C  is  a  sample  of  interstitial 
water  after  24  hours  exposure  to  a  sand  dollar ;  total  primary  amine  content  is  approximately 
875  picomoles.  His,  orn  and  gin  are  identified  by  Rf  values  from  other  standards. 

Values  for  oxygen  consumption  were  not  found  for  sand  dollars  in  the  literature. 
The  measurements  presented  in  this  study  are  intended  to  offer  an  approximate 
figure  for  oxygen  consumption  for  comparison  with  measured  rates  of  amino  acid 
uptake.  Two  small  animals  (1.8  and  1.9  cm  average  diameters)  consumed  4.6  and 
4.7  jul  Oo/(hr-cnr)  ;  two  large  animals  (6.15  and  5.75  cm  average  diameter)  con- 
sumed 5.0  and  2.8  /xl  Oo/(hr-cm2).  In  contrast  to  the  relative  constancy  of  oxygen 
consumption  expressed  per  unit  surface,  oxygen  consumption  per  unit  weight  de- 
creased rapidly  with  size  as  would  be  anticipated.  For  the  small  animals,  rates 
were  46.0  and  43.4  pi  O2/(g-hr)  ;  for  the  larger  animals,  they  were  13.2  and  15.1 
/A  O2/(g'hr).  Despite  the  small  sample  size,  it  seems  reasonable  to  accept  the 
average  figure  of  4.3  fj.1  O2/(hr-cm2)  as  an  estimate  of  typical  oxygen  consumption 
at  20°  C.  Isolated  portions  of  the  aboral  test  survived  well  in  aerated  sea  water 
for  two  to  three  days  at  20°  C  as  judged  by  general  appearance  and  activity  of 
pedicellariae.  Oxygen  consumption  of  two  such  portions  of  the  test  was  6.3  and 
3.7  ju,l  Oo/(hr-cm'-),  respectively.  The  subdermal  portion  of  the  test  was  cleaned  of 
adherent  tissues,  but  the  epithelium  contributed  to  oxygen  consumption  ;  however, 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER  343 

it  appears  that  the  epidermis  is  responsible  for  a  large  fraction  of  the  total  oxygen 
consumption  of  the  animals. 

DISCUSSION 

Simultaneous  measurement  of  influx  and  net  influx  of  seven  amino  acids  indi- 
cate that  neutral  amino  acids  (ala,  gly,  ser,  val)  are  removed  rapidly  from  dilute 
solution  in  ambient  sea  water  by  Dcndrastcr.  The  amino  acids  asp  and  lys  enter 
more  slowly;  glu  is  removed  from  solution  very  slowly,  if  at  all  (Table  I).  Entry 
rates  as  estimated  by  disappearance  of  14C-labeled  substrate  and  by  chemical  deter- 
mination of  total  primary  amine  remaining  in  solution  are  comparable  at  ambient 
concentrations  of  5  /AM  or  more  (Fig.  1).  Thus,  estimates  of  influx  (14C)  reflect 
net  influx  (primary  amine)  at  concentrations  which  are  normally  present  in  the 
habitat  of  the  organism. 

When  Dcndraster  is  placed  in  a  fixed  volume  of  sea  water,  an  efflux  of  primary 
amines  of  unknown  composition  occurs  until  an  apparent  steady  state  is  reached  at 
an  ambient  concentration  of  5-7  /*M.  Efflux  appears  to  be  slow  compared  with 
uptake  of  neutral  amino  acids.  Thus,  the  pattern  of  primary  amine  concentration 
in  the  medium  with  time  may  show  a  decrease  with  a  subsequent  increase  (Fig.  2). 

Dcndraster  is  capable  of  net  accumulation  from  solution  of  some  of  the  naturally 
occurring  primary  amines  found  in  the  interstitial  water  of  sediment  from  its  habitat. 
Rates  of  removal  are  approximately  half  the  rates  observed  for  neutral  amino  acids 
(Table  I)  when  expressed  in  comparable  units.  Two  explanations  for  this  lower 
rate  can  be  suggested.  First,  glu  is  present  in  interstitial  water  and  therefore  con- 
tributes to  total  primary  amine  but  is  relatively  unavailable  to  the  animal.  Secondly, 
15-30%  of  the  primary  amines  in  interstitial  water  are  not  retained  on  Dowex-50  in 
the  acid  phase  and  may  represent  material,  some  or  all  of  which  is  unavailable  for 
transepidermal  uptake. 

Comparison  of  the  amino  acids  present  in  interstitial  water  before  and  after  ex- 
posure to  Dcndraster  shows  a  change  in  total  primary  amines  and  a  change  in  pat- 
tern of  amino  acids  present  (Fig.  4).  The  changes  are  consistent  with  predictions 
based  on  experimental  results  with  single  amino  acids.  Thus,  neutral  amino  acids 
are  reduced,  while  glu  is  relatively  unchanged;  total  primary  amines  are  reduced 
to  a  stable  level  of  5-7  /X.M. 

As  noted,  a  portion  of  the  primary  amines  normally  present  in  interstitial  water 
does  not  appear  to  behave  as  typical  amino  acid.  Changes  in  the  contribution  of 
this  fraction  to  total  primary  amines  during  exposure  to  Dcndrastcr  were  not  deter- 
mined. Estimation  of  its  concentration  by  difference  before  and  after  passing 
through  a  Dowex-50  column  proved  to  be  unsatisfactory.  Presence  of  this  unknown 
material  also  prohibits  a  complete  description  of  the  primary  amines  which  appear 
in  the  medium  in  which  animals  are  incubated. 

Amino  acids  removed  from  solution  by  these  animals  apparently  enter  meta- 
bolic pathways.  In  general,  acidification  of  the  medium  leads  to  a  reduction  in 
measured  radioactivity  of  a  medium  sample  after  an  animal  has  been  exposed  to  a 
known  labelled  substance.  This  acid  volatile  radioactivity  may  be  evidence  for  the 
production  of  14COo,  commonly  found  in  experiments  of  this  kind  ( Stephens,  1972). 


344  STEPHENS,  VOLK,  WRIGHT,  AXU  BACKLUND 

The  presence  of  radioactivity  which  is  not  acid  volatile  and  which  is  not  ainino 
acid  at  the  end  of  incubation  experiments  may  he  evidence  of  the  presence  of 
labelled  metabolites  lost  from  the  animals.  These  have  not  been  identified  but  are 
not  primary  amines. 

The  failure  of  Dendraster  to  remove  gylcylglycine  from  dilute  solution  suggests 
that  if  epidermal  digestion  does  occur  in  these  animals,  it  is  a  slow  process  compared 
to  transepidermal  transport  of  amino  acids.  This  is  consistent  with  the  very  low 
protein  digestion  activities  reported  by  Pequignat  (1970),  but  other  pathways  of 
disappearance  of  glycylglycine  cannot  be  excluded  in  these  observations. 

Possible  bacterial  contributions  to  the  appearance  of  labeled,  acid  volatile  and 
acid  nonvolatile  metabolites  in  the  medium  and  to  the  disappearance  of  glycylglycine 
cannot  be  completely  excluded.  However,  bacterial  contributions  to  influx  and  net 
flux  measurements  are  certainly  small.  Animals  were  incubated  for  24  hours  in 
penicillin  (500,000  units/liter)  and  streptomycin  (200  nig/liter)  and  rates  of  influx 
and  net  flux  of  lysine  and  serine  compared  to  unincubated  controls.  No  difference 
was  observed.  Such  incubation  would  not  inactivate  all  possible  microbial  con- 
taminants, but  one  would  anticipate  some  effect  on  rates  if  microbial  activity  plays 
a  substantial  role  in  these  observations.  Failure  to  observe  influx  of  glutamate  iu 
Dendraster  also  suggests  that  the  animal  is  the  principal  agent ;  there  is  no  reason 
to  expect  that  glutamate  would  not  be  metabolized  as  well  as  other  amino  acid  sub- 
strates by  a  contaminant  microbial  population. 

The  potential  contribution  of  transepidermal  transport  to  the  animals  can  be 
estimated  by  comparing  rates  of  influx  to  an  estimate  of  reduced  carbon  required  to 
support  oxidative  metabolism.  An  approximate  conversion  factor  to  equate  oxygen 
consumption  with  complete  oxidation  of  a  mixture  of  amino  acids  (1  ml  Oo  =:  1  mg 
amino  acid)  and  an  average  molecular  weight  for  amino  acids  of  100  can  be  used. 
Then,  the  average  oxygen  consumption  of  4.3  jA  (^/(hr-cni2)  is  equivalent  to  43 
mnoles  amino  acid/(hr-cnr).  The  average  influx  of  naturally  occurring  amines 
from  interstitial  water  (Table  I )  is  24.9  mnoles/ (hr- cm2),  a  contribution  of  58%  of 
the  material  required  to  support  oxygen  consumption.  This  estimate  is  probably 
based  on  an  overly  conservative  figure  for  the  level  of  naturally  occurring  amines 
in  the  sediment.  Only  two  of  the  fifteen  cores  analyzed  from  the  habitat  showed 
less  than  50  /mioles  primary  amines  (17,36  /Ainoles  )  in  the  0-3  cm  zone  of  the  sedi- 
ment. The  average  was  115  /J.M.  Since  the  Kt  for  influx  of  gly  was  measured  as  74 
/AM  (  Fig.  3),  the  presence  of  levels  of  primary  amines  in  interstital  water  greater 
than  the  20  /AM  used  for  this  estimate  would  certainly  lead  to  greater  influx  rates 
and  an  increased  contribution  to  carbon  requirements.  In  fact,  it  can  be  concluded 
that  if  the  surface  of  Dendraster  is  exposed  to  levels  of  primary  amines  measured  in 
14  or  our  15  samples  (>35  /AM),  influx  is  sufficient  to  account  for  oxygen  con- 
sumption. 

This  discussion  assumes  that  the  bulk  concentration  of  primary  amines  measured 
in  the  interstitial  water  of  the  sediment  is  a  measure  of  concentrations  available  at 
the  surface  of  the  animal.  Stephens  (1975)  has  reported  increased  primary  amines 
in  interstitial  water  as  a  result  of  irrigation  by  the  annelid  infauna.  This  may  also 
be  true  for  Dendraster.  Alternatively,  renewal  of  primary  amines  at  the  surface 
may  be  dependent  on  bulk  flow  of  interstitial  water  and  diffusion.  Until  this  ques- 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER  345 

tion  can  be  investigated,  it  should  be  re-emphasized  that  the  conclusions  of  the 
preceding  paragraph  should  be  phrased  conditionally. 

Timko  (1976)  describes  suspension  feeding  in  Dcndraster.  When  behaving  in 
this  fashion,  about  one-third  of  the  anterior  portion  of  the  test  is  embedded  in  the 
sediment  substrate.  Clearly,  only  a  portion  of  the  test  would  be  in  contact  with 
interstitial  water  of  the  sediment  in  this  feeding  mode.  Dcndraster  also  behaves  as  a 
prone  deposit  feeder  according  to  Timko  and  other  authors.  In  this  feeding  mode, 
the  animal  is  often  below  the  sediment  surface  and  is  fully  exposed  to  interstitial 
water.  Animals  in  both  the  inclined  suspension  feeding  and  prone  deposit  feeding 
orientation  were  observed  in  the  shallow  water  population  at  Point  Mugu  Naval 
Base. 

Timko  (1976)  concludes  that  Dcndraster  c.rccntriciis  is  primarily  a  suspension 
feeder.  However,  Chia  (1969)  reports  that  all  the  individuals  in  a  population 
from  Puget  Sound,  Washington,  were  completely  buried  at  low  tide.  We  suggest 
that  Dcndraster  can  supplement  both  suspension  feeding  and  deposit  feeding  by 
influx  of  amino  acids  into  the  epidermis.  This  supplement  would  be  small  when 
the  animals  are  behaving  as  inclined  suspension  feeders  but  would  be  large  during 
deposit  feeding.  In  our  experiments,  transepidermal  influx  of  amino  acids  would 
support  energy  metabolism  at  ambient  levels  of  primary  amines  greater  than  35  /AM  ; 
our  measurements  indicate  these  are  realistic  levels  for  prone  deposit  feeding  animals 
buried  in  the  superficial  layers  of  the  sediment. 

Our  data  suggest  that  animals  might  indeed  survive  without  taking  in  and 
digesting  food,  provided  Dendraster  has  pathways  for  translocating  nutrients  from 
the  epidermis  to  deeper  tissues.  However,  it  is  more  likely  that  transepidermal  up- 
take of  small  organic  compounds  may  contribute  to  the  sustenance  of  the  epidermis. 
If  there  is  a  barrier  to  translocation  of  nutrients  in  Dcndraster  comparable  to  that 
reported  for  other  asteroids  and  echinoids,  direct  uptake  of  nutrients  from  the  en- 
vironment may  play  a  large  role  in  the  nutrition  of  pedicellariae,  spicules,  podia  and 
other  epidermal  structures.  Our  data  suggest  that  the  oxidative  requirements  of 
the  epidermis  represent  a  large  fraction  of  the  total  requirements  of  the  animal. 
However,  levels  of  ambient  primary  amines  (>  35  /AM)  adequate  to  support  total 
oxidative  metabolism  are  a  fortiori  adequate  for  the  epidermal  fraction  thereof. 

SUMMARY 

1.  Influx  of  amino  acids  from  dilute  solution  into  the  sand  dollar,  Dendraster, 
was  measured  by  following  the  disappearance  of  radioactivity  in  the  medium  supply- 
ing known  labeled  substrates.    Net  flux  was  monitored  simultaneously  by  following 
the  decrease  in  primary  amines  in  the  medium  fluorometrically.     Rates  of  influx 
and  net  flux  correspond  closely  at  ambient  concentrations  greater  than  5  /AM. 

2.  Dcndraster  is  capable  of  net  accumulation  of  some  of  the  primary  amines 
normally  found  in  the  interstitial  water  of  its  sediment  habitat. 

3.  A  sensitive  method  for  location  of  amino  acids  on  thin  layer  chromatograms 
is  described.     Comparison  of  interstitial  water  before  and  after  exposure  to  Den- 
draster shows  a  changed  pattern  of  amino  acids,  as  well  as  a  decrease  in  total  amino 
acids,  which  is  consistent  with  measurements  of  rates  of  influx   with  single  sub- 
strates. 


346  STEPHENS,  YOLK,   WRIGHT,  AND   BACKLUND 

4.  Comparison  of  rates  of  influx  of  naturally  occurring  primary  amines  with  the 
metaholic  requirements  of  animals  as  estimated  from  their  oxygen  consumption  indi- 
cates that  Dendrastcr  can  acquire  sufficient  reduced  carbon  to  account  for  its  oxida- 
tive  needs  if  its  surface  is  exposed  to  naturally  occurring  primary  amines  at  con- 
centrations greater  than  or  equal  to  35  /*M. 

5.  Primary  amines  in  the  interstitial  water  of  sediment  in  the  immediate  vicinity 
of  a  shallow  water  population  of  Dendrastcr  range  in  concentration  from  17  to  244 

(115  ±60 /AM). 

6.  Dendrastcr  lives  in  an  environment  which  is  relatively  rich  in  amino  acids, 
and  it  possesses  a  transport  system  which  can  accumulate  these  compounds  at  rates 
sufficient  to  provide  a  significant  supplement  to  other  forms  of  feeding.     These 
findings  support  the  hypothesis  that  sustenance  of  epidermal  structures  of  echinoids 
and  asteroids  may  be  relatively  independent  of  translocation  of  nutrients  from  the 
digestive  organs  and  may  be  based  primarily  on  transepidermal  influx  of  nutrients 
from  the  medium. 


LITERATURE  CITED 

AHEARN,  G.  A.,  AND  S.  J.  TOWNSLEY,  1975.  Integumentary  amino  acid  transport  and  meta- 
bolism in  the  apodous  sea  cucumber,  Chiridota  rif/ida.  J.  E.vp.  Bio].,  62:  733-752. 

CAVANAUGH,  G.  M.  (Ed.),  1956.  Formulae  and  Methods,  IV,  of  the  Marine  Biological  Labora- 
tory Chemical  Room.  Marine  Biological  Laboratory,  Woods  Hole,  Massachusetts,  61  pp. 

CHIA,  F.  S.,  1969.  Some  observations  on  the  locomotion  and  feeding  of  the  sand  dollar, 
Dendraster  e.vcentricits.  J.  E.\-p.  Mar.  Biol.  Ecol.,  3:  162-170. 

CLARK,  M.  E.,  1968.  Simple,  rapid  quantitative  determination  of  amino  acids  by  thin-layer 
chromatography.  Analyst,  93  :  810-816. 

CLAKK,  M.  E.,  1969.  Dissolved  free  amino  acids  in  sea  water  and  their  contribution  to  the 
nutrition  of  sea  urchins.  Pages  70-93  in  Annual  Report  Kelp  Habitat  Improvement 
Projects,  1968-1969.  W.  M.  Keck  Lab.  of  Environmental  Health  Engineering,  Cali- 
fornia Institute  of  Technology,  Pasadena. 

CROWE,  J.  H.,  K.  A.  DICKSON,  J.  L.  OTTO,  R.  D.  COLON,  AND  K.  K.  FARLEY,  1977.  Uptake 
of  amino  acids  by  the  mussel,  Modiolus  demissus.  J.  E.vp.  Zool.,  202  :  323-332. 

DIXIT,  D.  B.,  1973.  Uptake  of  amino  acids  and  development  in  the  sea  urchin,  Strongylo- 
centrotus  purpuratux.  I'h.D.  dissertation,  University  of  California,  Irvine,  117  pp. 

FERGUSON,  J.  C.,  1967.  An  autoradiographic  study  of  the  utilization  of  free  exogenous  amino 
acids  by  starfishes.  Biol.  Bull.,  133:  317-329. 

FERGUSON,  J.  C.,  1970.  An  autoradiographic  study  of  the  translocation  and  utilization  of  amino 
acids  by  starfish.  Biol.  Bull.,  138  :  14-25. 

FERGUSON,  J.  C.,  1971.  LTptake  and  release  of  free  amino  acids  by  starfishes.  Biol.  Bull.,  141  : 
122-129. 

FONTAINE,  A.  R.,  AND  F.  CHIA,  1968.  Echinoderms :  an  autoradiographic  study  of  assimilation 
of  dissolved  organic  molecules.  Science,  161:  1153-1155. 

MENDEZ,  E.,  AND  J.  G.  GAVILANES,  1976.  Fluorometric  detection  of  peptides  after  column 
chromatography  or  on  paper:  o-phthalaldehyde  and  fluorescamine.  Anal.  Biochem., 
72  :  473-479. 

NORTH,  B.  B.,  1975.  Primary  amines  in  California  coastal  waters:  utilization  by  phyto- 
plankton.  Limnol.  (h'cuno/ir..  20:  20-27. 

PEARSE,  J.  S.,  AND  V.  B.  PEARSE,  1973.  Removal  of  glycine  from  solution  by  the  sea  urchin. 
Strongyloccntrotus  purpiiratus.  Mar.  Biol.,  19:  281-284. 

PEQUIGNAT,  E.,  1969.  Sur  1'absorption  et  1'utilisation  de  molecules  dissoutes  ainsi  que  des 
particules  en  suspension  par  les  oursins  reguliers  et  irreguliers.  C.  R.  Seances  Soc. 
Biol.Fil.,  163:  101-104. 

PEQUIGNAT,  E.,  1970.     Biologic  des  Bchinocardium  cut-datum  (Pennant)   de  la  Baie  de  Seine. 


UPTAKE  OF  AMINO  ACIDS  BY  DENDRASTER  347 

Nouvelles  recherches  stir  la  digestion  et  1'absorption  cutanees  chex  les  Echinides  et  les 
Stellerides.  Fonna  Functio,  2 :  121-168. 

PEQUIGNAT,  E.,  AND  J.  PUJOL,  1968.  Absorption  cutanee  de  3H-proline  a  tres  faible  concen- 
tration et  son  incorporation  dans  le  collagene  chez  Psammechinus  miliaris.  Bull.  Soc. 
Linn.  Xormandic  lOc  Scric,  9:  209-219. 

SCHILTZ,  E.,  K.  D.  SCHNACKERZ,  AM)  R.  \V.  GRACY,  1977.  Comparison  of  ninhydrin,  fluores- 
camine,  and  o-phthaldialdehyde  for  the  detection  of  amino  acids  and  peptides  and  their 
effects  on  the  recovery  and  composition  of  peptides  from  thin-layer  fingerprints.  Anal. 
Biochcm.,  79:  33-41. 

STEPHENS,  G.  C.,  1972.  Amino  acid  accumulation  and  assimilation  in  marine  organisms.  Pages 
155-184  in  T.  W.  Campbell  and  L.  Goldstein,  Eds.,  Nitrogen  metabolism  and  the  en- 
vironment. Academic  Press,  New  York. 

STEPHENS,  G.  C.,  1975.  Uptake  of  naturally  occurring  primary  amines  by  marine  annelids. 
Biol.  Bull.,  149:  397-407. 

STEPHENS,  G.  C.,  AND  R.  A.  SCHINSKE,  1961.  Uptake  of  amino  acids  by  marine  invertebrates. 
Liinnol.  Occanof/r.,  6:  175-181. 

STEPHENS,  G.  C.,  AND  R.  A.  VIRKAR,  1966.  Uptake  of  organic  material  by  aquatic  inverte- 
brates. IV.  The  influence  of  salinity  on  the  uptake  of  amino  acids  by  the  brittle  star, 
Ophiactis  arcnosa.  Biol.  Bull.,  131:  172-185. 

TIMKO,  P.  L.,  1976.  Sand  dollars  as  suspension  feeders:  a  new  description  of  feeding  in  Dcn- 
dr aster  cxccntricus.  Biol.  Bull.,  151 :  247-259. 


Reference:   liiol.  />'»//..  154 :  348-359.    (April,  1978) 


OCCURRENCE  AND  GROUP  ORGANIZATION  OF  ATLANTIC 

BOTTLENOSE  PORPOISES  (TURSIOPS  TRUNCATUS) 

IN  AN  ARGENTINE  BAY 

BERND  WtiRSIG 

Prayram  for  Neurobiology  mid  Behavior,  State  University  of  New  York, 
Stony  lironk,  New  York  11794 

While  the  social  behavior  of  many  terrestrial  mammals  has  been  well  described 
(see  Wilson,  1975,  pages  456-540,  for  a  review),  much  less  is  known  about  the 
social  organization  of  the  several  species  of  porpoises  that  inhabit  all  oceans  of  the 
world  (Norris  and  Dohl,  1978a,  provide  a  review).  This  lack  of  information  re- 
sults from  the  difficulty  of  remaining  with  a  group  of  porpoises  in  the  open  ocean 
long  enough  to  observe  the  details  of  porpoise  behavior,  and  from  the  interference 
with  the  animals'  behavior  that  a  boat  causes. 

There  are  places  where  porpoises  come  close  enough  to  shore  to  make  observa- 
tions from  land  feasible  (Mitchell,  1975).  Saayman,  Bower,  and  Tayler  (1972) 
described  the  activity  cycles  and  movements  of  Indian  Ocean  bottlenose  porpoises 
and  Indopacific  humpback  porpoises  by  observations  from  South  African  cliffs ;  and 
Norris  and  Dohl  (Norris,  1974;  Norris  and  Dohl,  1978b)  made  similar  observa- 
tions on  Hawaiian  spinner  porpoises  from  shore  vantage  points.  In  the  present 
study,  Atlantic  bottlenose  porpoises  were  observed  from  a  45-meter  cliff  located  on 
the  coast  of  south-Argentina.  The  purpose  of  this  paper  is  to  describe  the  seasonal 
pattern  of  occurrence,  group  stability,  surfacing  associations,  and  calving  seasonally 
of  these  animals.  These  data  represent  a  first  step  in  understanding  the  social  be- 
havior of  the  bottlenose  porpoise.  Some  information  on  the  group  stability  of  this 
porpoise  population  and  the  photographic  technique  used  to  gather  these  data  has 
been  presented  elsewhere  (Wiirsig  and  Wiirsig,  1977).  The  present  paper  is  a 
more  complete  treatment  of  this  material. 

MATERIALS  AND  METHODS 

During  a  21 -month  stay,  from  July  1974  through  March  1976,  at  Golfo  San 
Jose  (42°  23'  S,  64°  03'  W),  bottlenose  porpoises,  Titrsiops  tniticafus,  were  ob- 
served as  they  periodically  passed  within  one  kilometer  of  a  shore  observation  point 
(camp).  To  investigate  the  group  composition  and  stability  of  this  population,  por- 
poises were  identified  by  photographing  the  natural  markings  on  the  trailing  edges 
of  their  dorsal  fins  (see  Wiirsig  and  Wiirsig,  1977).  Observations  lasted  from  ten 
minutes  to  several  hours,  depending  on  the  length  of  time  that  the  porpoises  stayed 
near  shore.  It  was  assumed  that  all  porpoises  were  photographed  when  each  animal 
was  identified  at  least  four  times  within  the  record  of  one  photographic  observation 
session. 

In  the  present  paper,  (jronp  refers  to  53  individually  identified  bottlenose  por- 
poises which  passed  through  the  study  area  during  a  21 -month  period.  This  group 

348 


PORPOISE  GROUP  ORGANIZATION 


is  part  of  a  larger  population  of  unknown  size.  Subgroup  refers  to  those  animals  of 
the  group  which  passed  by  shore  at  any  one  time. 

To  assess  surfacing  associations  of  animals,  a  motordrive  Nikon  camera  was 
used.  This  provided  data  not  only  on  which  individuals  were  present,  but  also  on 
their  dive  times  and  on  which  individuals  surfaced  together.  To  collect  this  informa- 
tion, a  photograph  was  taken  each  time  one  or  more  animals  surfaced  (up  to  1.5 
frames/sec  could  be  taken).  The  camera  clicks  were  recorded  on  magnetic  tape, 
and  comparison  of  times  between  photographs  and  the  individuals  recognized  in 
those  photographs  provided  individual  dive  times  and  a  measure  of  whether  any 
animals  tended  to  surface  at  nearly  the  same  time.  The  technique  can  be  profitably 
used  when  the  animals  are  close  enough  to  the  camera  to  allow  for  recognition  of 
all  individuals  as  they  surface.  The  use  of  a  35  mm  motordrive  camera  provided 
large  negatives  with  the  detail  necessary  for  recognizing  individual  animals;  yet  the 
rate  of  picture  taking  was  sufficient  to  photograph  all  animals  as  they  surfaced.  A 
cine-camera  technique  for  determining  group  size,  deployment,  and  speed  (but  not 
recognition  of  individuals)  was  described  by  Tayler  and  Saayman  (1972a). 

Seasonal  occurrence  patterns  were  analyzed  using  analysis  of  variance  (Sokal 


JUL    AUG     SEP    OCT     NOV    DEC     JAN     FEB    MAR     APR    MAY     JUN 

MONTHS 

FIGURE  1.  The  fraction  of  the  possible  days  each  month  on  which  hottlenose  porpoises  were 
sighted.  The  Y-axis  represents  the  ratio  of  the  number  of  days  on  which  porpoises  were  sighted 
divided  by  the  number  of  days  each  month  with  winds  less  than  20  km/hr.  July,  November, 
and  March  were  months  of  maximum  porpoise  sightings ;  in  September,  February,  and  May 
they  were  sighted  significantly  less  (P  <  0.001,  analysis  of  variance,  Sokal  and  Rohlf,  1969, 
pages  204-249,  and  Rohlf  and  Sokal,  1969,  pages  168-197). 


350 


BERND  WORSIG 


and  Rohlf,  1969;  Rolilf  and  Sokal,  1969),  surfacing  associations  were  tested  with  a 
sampled  randomization  test  (Sokal  and  Knlilf.  1909),  and  significance  of  calving 
seasonally  was  obtained  with  the  Raleigh  test,  using  the  procedure  described  by 
Greenwood  and  Durand  (1955). 

RESULTS 

Seasonal  occurrence  pattern 

On  days  with  winds  greater  than  20  km/hr  it  was  difficult  to  see  or  photograph 
porpoises.  Of  the  433  days  with  winds  less  than  20  km/hr,  bottlenose  porpoises 
were  seen  on  191  days,  or  44%  of  the  days  on  which  observations  were  made.  The 
number  of  days  on  which  porpoises  were  sighted  varied  greatly  from  month  to 
month.  As  Figure  1  shows,  porpoises  were  seen  near  shore  during  20  of  the  21 
months  studied;  August,  1974,  was  the  only  month  without  sightings.  But  the 
number  of  days  on  which  they  were  sighted  varied  greatly  from  month  to  month  : 
there  was  a  peak  of  abundance  about  every  four  months,  and  this  pattern  was 
similar  for  the  two  years. 

Subgroup  composition  and  stability 

Bottlenose  porpoise  subgroups  were  photographed  on  approximately  150  of  the 
191  sighting  days.  There  are  35  days  on  which  all  individuals  in  a  subgroup  were 
photographed  at  least  four  times.  Figure  3  shows  which  of  the  53  known  indi- 
viduals were  present  on  each  of  these  35  days.  Only  one  subgroup  was  sighted 
during  any  one  day.  The  number  of  individuals  in  a  subgroup  varied  from  eight  to 
22,  with  a  mean  of  14.9  (s.d.  =  3.28,  see  Figure  2). 


o 

OCCURRENCES 
n  CD  ->i  oo  cj 

1  — 

^i 

L_ 
O      A. 

H 

CC 

1   r    1             "^ 

UJ       O 

m 

^         0 

Z)       <- 
1 

1 

n 

NUMBER    OF    INDIVIDUALS    PER    SUBGROUP 

FIGURE  2.     A  histogram  of  subgroup  sizes  during  35  days  on  which  all  individuals  present 
were  recognized. 


PORPOISE  GROUP  ORGANIZATION 


351 


COMPLETE    SIGHTING    DATES 
1975 


1976 


o 
O 
o 

LjJ 

cr 


ojrooiOoji^oo      ^ro^^^ro^rcn  — ,  |{2  59  £]  9.  ~  ®  —         t 


FIGURE  3.  Specimens  of  Tursiops  recognized  during  a  21-month  continuous  study  period. 
Shaded  blocks  represent  the  presence  of  individuals  during  35  complete  sighting  days  (see  text). 
The  last  three  shaded  columns  on  the  right  represent  additional  sightings  of  animals  first  at 
300  kilometers  removed  from  camp,  and,  secondly  and  thirdly,  animals  found  near  camp  Decem- 
ber, 1976. 


As  Figure  3  makes  clear,  this  variation  in  subgroup  size  was  the  result  of  a 
continual  flux  of  animals  leaving  and  joining  the  subgroups.  Nevertheless,  some 
animals,  namely  #1-5,  may  be  termed  "core"  animals  since  they  were  present 
throughout  the  study  period.  Animals  #7-12  were  present  during  the  first  ten 
months  but  subsequently  disappeared  during  the  month  when  animals  #14-18  first 
appeared.  In  addition  to  these  major  changes  in  subgroup  composition,  other 


352 


BKKXI)    \\  rkSHi 


individuals  appeared  and  disappeared  together.  For  example,  #2,  #9,  #13,  and 
#22  disappeared  between  October  13,  1974,  and  November  21,  1974,  while  #24 
appeared  during  their  absence.  At  that  time  as  well,  four  others  (#25  and  its  calf- 
#36,  #35,  and  #37)  appeared  and  stayed  until  January  5,  1975.  From  January 
5,  1975,  to  March  19,  1975,  five  animals  (#4  and  its  ca"lf-#5,  #9,  #12,  and  #13) 
disappeared  while  once  again  an  individual  (#32)  appeared  in  their  absence.  AYhen 
these  five  reappeared  and  #32  disappeared  March  19,  two  new  animals  appeared 
(#28  and  #51).  Porpoises  #51,  #28,  and  #13  disappeared  again  by  April  25, 
1975.  Porpoise  #6,  which  was  absent  from  March  19  to  October  23,  1975,  re- 
appeared with  the  large  shift  in  individuals  first  documented  on  that  date.  A 
monthly  summary  of  the  presence  and  absence  of  22  of  the  animals  described  above 
is  shown  in  Figure  4.  The  appearance  and  disappearance  of  other  animals  was 
apparently  not  related  to  the  presence  or  absence  of  conspecifics. 

Porpoises  #7-12,  which  disappeared  from  camp  in  September  1975,  were 
spotted  in  March,  1976,  over  300  kilometers  from  the  study  site.  On  December  1 1 
and  29,  1976,  after  nine  months  of  no  observations,  a  spot-check  of  porpoises  near 


1974 


1975 

A 


1976 


OCT  NOV  DEC  JAN  FEB  MAR  APR  MAY  JUN  JUL  AUG  SEP  OCT  NOV  DEC  JAN  FEB  MAR 


J 


J 


r~~i     t '.  '.  V^^SNS^ ^>^^\\^^^y^->- 


RESIGHTED   MARCH,  76 
|>OVER  300km 
DISTANT  FROM  CAMP 


t  J 


2 

3 

4 

5 

6 

1 

8 

9 

10 

II 

12 

14 

15 

16 

17 

18 

32 

25 

36 

37 

35 


FIGURE  4.  A  month-by-month  summary  of  the  presence  of  22  individuals  which  show 
interpersonal  relationships  in  presence  or  absence  within  a  subgroup.  Differences  in  the  sum- 
mary and  Figure  3  are  due  to  consideration  here  of  all  sighting  days,  while  Figure  3  represents 
only  the  35  complete  sighting  days  when  all  individuals  were  recognized. 


PORPOISE  GROUP  ORGANIZATION  353 

TABLE  I 

Pairs  of  porpoises  surfacing  within  three  seconds  of  each  other  on  two  days  in  January,  1976  \_(ns) 
=  not  surfacing  significantly  together;  pro/ia/nlity  of  random  on  nrrence,  P  <  0.025*,  or  P  <  0.001**, 
sampled  randomization  test~\. 

January  18,  1976  January  29,  1976 


#4-#5** 

#3-#52*  #3-#52(ns) 

#2-#6*  (only  #6  present) 

#2-#52*  (only  #52  present) 

#27-#6*  #27-#6  (ns) 

#6-#18*  #6-#18(ns) 

#4-#27  (ns)  #4-#27* 

(only  #52  present)  #52-#50** 


camp  showed  that  four  of  these  six  animals  were  again  in  the  original  study  area. 
At  that  time,  12  of  the  53  known  animals  were  present  but  no  new  animals  were 
recorded.  Interestingly,  of  the  core  unit  of  five  animals,  only  #4  and  #5  were 
present,  suggesting  that  even  the  "core"  may  not  be  stable  over  a  larger  time  period. 

Surfacing  associations  and  dive  times 

To  determine  whether  there  is  any  association  between  animals  breathing 
together  which  might  indicate  a  social  bond,  the  surfacing  pattern  of  porpoises  was 
examined  for  two  days  in  January,  1976.  As  Table  I  shows,  while  adult-adult 
surfacing  associations  occurred,  they  were  not  the  same  associations  during  the  two 
sampled  days,  and  they  were  never  of  greater  significance  than  at  the  P  <  0.025 
level  (sampled  randomization  test,  Sokal  and  Rohlf,  1969,  p.  633).  Adult-calf 
associations,  however,  were  observed  for  #4  and  its  calf  (#5)  on  both  days,  and 
were  highly  significant  (P  <  0.001).  Porpoise  #52  and  its  calf  (#50)  were  pres- 
ent on  only  one  of  the  two  sampled  days  (January  29,  1976),  and  their  association 
at  that  time  was  also  highly  significant  (P  <  0.001).  Except  for  adult-calf  surfac- 
ing associations  there  was  no  detectable  stable  relationship  in  animals  surfacing 
together  (Table  I)  and  animals  appearing  together  in  subgroups  on  different  days 
(Fig.  3). 

Because  all  individuals  were  recognized  as  they  surfaced  during  the  two  sampled 
sessions,  dive  time  data  were  accumulated  as  well.  Individuals  dove  for  a  mean  of 
21.8  seconds,  with  little  fluctuation  from  this  mean  (s.d.  ==  3.01).  The  small  calf 
present  on  January  29,  1976,  surfaced  slightly  more  frequently  than  the  overall 
mean  (#50  --  17.8  sec/dive)  and  slightly  more  frequently  than  its  presumed 
mother  (#52  —  21.3  sec/dive). 

Calving  seasonality 

Five  calves  were  observed  throughout  the  study,  and  two  were  reported  after 
termination  of  the  study  (Table  II).  The  mean  number  of  calves  per  subgroup  was 
1.5  for  the  35  sightings  of  completely-known  groupings  to  pass  by  camp.  On  the 
average,  10%  of  a  subgroup  was  composed  of  calves.  Calves  were  born  during 


354  BERND  WURSK; 

TABLE  1 1 

The  estimated  birthdalcs  of  seven  calves,  arranged  by  month.  The  two  un-numbered  calves  were  re- 
ported first  seen  in  the  months  shown,  after  termination  of  the  21-month  study.  Note  the  absence  of 
births  May  through  October. 

Months  Estimated  dates  of  calf  births 

January  #21  (1975);  #50  (1976) 

February  none 

March   '  #53  (1976);  no  number  (1977) 

April  #5  (1974);  no  number  (1977) 

May  none 

June  none 

July  none 

August  none 

September  none 

October  none 

November  #36  (1974) 

December  none 


spring,  summer,  and  fall,  with  no  births  during  winter  (June- September).  The 
births  of  seven  calves  during  the  six  month  period  from  November  through  April 
(see  Table  II)  are  nonrandomly  clumped  toward  those  months  (P  <  0.02,  Raleigh 
test,  Greenwood  and  Durand,  1955).  There  was  no  evidence  of  births  for  May 
through  October. 

DISCUSSION 

Recognized  individuals  of  bottlenose  porpoises  were  found  in  the  study  area 
throughout  the  year  and  thus  did  not  migrate  with  the  changing  seasons.  True 
(1891)  reported  that  specimens  of  Tiirsiops  off  Cape  Hatteras,  North  Carolina, 
move  toward  the  north  in  the  spring  and  return  south  in  the  fall  (see  also  Mead, 
1975).  Gunter  (1942),  however,  reports  that  specimens  of  Tursiops  in  Texas 
waters  show  no  seasonal  migration;  while  Caldwell  and  Caldwell  (1972)  and 
Odell  (1975)  suggest  a  possible  seasonally-related  movement  of  this  species  off  the 
southern  tip  of  Florida.  Apparently  some  bottlenose  porpoise  populations  migrate 
and  others,  at  least  at  times,  do  not.  It  is  likely  that  the  porpoises  go  where  they 
can  find  food,  as  has  been  indicated  for  other  species  (Evans,  1971 ;  Norris  and 
Dohl,  1978a). 

Although  there  was  no  evidence  for  seasonal  migration  in  the  Argentine  study 
area,  there  was  a  four  months  cycle  in  the  number  of  times  that  porpoises  were 
sighted.  Lows  occurred  in  August-September,  January-February,  and  May-June; 
and  highs  in  July,  November,  and  March.  The  near-shore  surface  water  tem- 
perature in  the  study  area  varies  from  a  July- August  low  of  approximately  10.5°  C 
to  a  January-March  high  of  17  to  18°  C  (personal  observation).  Since  lows"  and 
highs  of  bottlenose  porpoise  presence  occurred  during  both  the  low  and  high  tem- 
perature periods  of  the  year,  as  well  as  during  intermediate  water  temperatures, 
occurrence  pattern  of  this  population  does  not  appear  temperature-dependent.  In 
other  areas  of  the  world,  this  species  is  also  found  over  a  wide  water  temperature 


PORPOISE  GROUP  ORGANIZATION  355 

range,  from  approximately  8°  C  to  30°  C  (Van  Bree,  University  of  Amsterdam, 
personal  communication) . 

Subgroups  which  were  seen  from  shore  during  the  21-month  study  varied  in 
size  and  in  composition  of  individuals  from  sighting  to  sighting.  Nevertheless,  five 
animals  of  a  recognized  group  of  53  animals  were  consistently  present  when  a  sub- 
group was  sighted.  These  five  individuals  were  composed  of  a  large  adult  (#1  ), 
two  smaller  adults  (#2  and  #3),  and  an  adult  (#4  )  with  its  calf  (#5).  Judging 
by  size,  it  is  possible  that  the  large  adult,  #1,  was  a  male  and  the  other  adults  were 
females.  This  kind  of  association  has  been  described  for  bottlenose  porpoises  in 
captivity  (Tavolga,  1966;  Caldwell  and  Caldwell.  1972),  as  well  as  in  the  wild 
(Caldwell  and  Caldwell,  1972;  Irvine,  University  of  Florida,  personal  communica- 
tion). In  the  present  study  this  supposition  rests  only  on  relative  size  of  individuals 
and  not  on  known  sex. 

In  addition  to  the  four  adults  and  one  calf  which  were  consistently  present, 
six  individuals  (#7-12)  were  present  until  September,  1975;  and  five  individuals 
(#  14—18)  were  present  from  September,  1975,  to  the  end  of  the  study.  These  two 
stable  groupings  were  composed  of  all  adults,  with  no  calves  or  juveniles  present. 
Caldwell  and  Caldwell  (1972)  hypothesized  that  such  units  may  be  composed  of 
nonbreeding  population  members.  They  may  travel  together  as  do,  for  example, 
bachelor  herds  of  elephants  (Douglas-Hamilton  and  Douglas-Hamilton,  1975),  but 
for  the  present  population  this  can  only  be  taken  as  a  suggestion,  in  need  of  further 
study. 

Perhaps  most  interesting  in  the  present  study  was  the  apparent  fluidity  with 
which  many  individuals  appeared  and  disappeared,  causing  a  constant  fluctuation  in 
subgroup  size  and  composition.  These  individuals  were  composed  of  adults  of 
varying  sizes,  and  of  calves  and  juveniles.  A  similar  situation  in  group  size 
fluctuation  exists  in  the  Hawaiian  spinner  porpoise,  Stcnclla  longirostris  (Xorris 
and  Dohl,  1978b).  A  possibly  similar  system  in  Florida  bottlenose  porpoises  has 
been  observed  recently  by  Wells  and  Irvine  (University  of  Florida,  personal 
communication)  and  may  be  found  to  be  quite  common  in  coastal  porpoise  species 
as  further  population  studies  are  made. 

Such  a  fluidity  in  stucture  surpasses  the  individual  interchanges  between  known 
"open"  groups  of  most  terrestrial  mammals  (Wilson.  1975,  pages  456-546).  To 
conform  to  the  standard  notion  of  groups  as  relatively  stable  units,  the  53  known 
individuals  of  the  present  population  have  been  labeled  yroitp,  while  the  units  that 
periodically  came  by  shore,  consisting  of  8  to  22  animals,  have  been  termed  sub- 
groups (\Viirsig  and  W'iirsig,  1977).  These  flexible  subgroups  appear  similar  to 
the  casual  units  found  in  chimpanzee  (Pan  troglodytes]  society  (Goodall,  1965  ; 
Reynolds  and  Reynolds,  1965;  Hall.  1968;  Xishida,  "l968;  Sugiyama,  1968).  The 
possibility  of  similarity  in  group  organization  between  Tursiops  and  Pan  was  first 
suggested  by  Tayler  and  Saayman  (1972b).  In  chimpanzee  society,  it  appears  that 
this  constant  fluctuation  in  subgroup  size  is  in  direct  response  to  irregular  and 
patchy  food  availability,  with  small  units  when  food  is  being  sought  and  larger  units 
in  areas  of  greater  food  abundance  (Reynolds,  1965).  It  is  suggested  that  food 
availability  may  also  be  a  primary  determinant  of  subgroup  size  and  stability  in  the 
present  bottlenose  porpoise  population. 


356  RERND  WORSIG 

Because  this  sludv  relied  on  sightings  at  a  discrete  point  along  tlie  shoreline, 
little  information  about  the  group's  total  range  was  gathered.  At  least  at  times 
some  of  the  animals  travelled  unexpectedly  long  distances,  however.  Six  individuals 
were  identified  in  a  bay  south  of  Golfo  San  Jose,  separated  from  camp  by  over  300 
kilometers,  and  nine  months  later  four  of  these  animals  were  again  found  near  camp. 
Either  the  normal  range  of  this  population  extends  over  so  extensive  a  distance, 
or  the  individuals  so  observed  had  crossed  into  the  new  area.  Similar  distances 
travelled  have  been  reported  for  pelagic  porpoises  (Perrin,  1975;  Evans,  1974),  but 
not  for  nonmigrating  near-shore  species. 

Although  some  adult  porpoises  showed  a  tendency  to  surface  together,  this  is 
not  a  long-term  relationship  and  may  shift  from  day  to  day.  Just  as  terrestrial 
animals  which  have  formed  close  social  bonds  do  not  in  general  exhibit  synchronized 
breathing,  porpoises  not  surfacing  together  may  still  be  closely  associated.  How- 
ever, porpoises,  unlike  terrestrial  animals,  must  leave  their  underwater  positions 
while  surfacing  to  take  a  breath.  As  a  result,  very  close  animal  associations  may 
be  reflected  in  the  breathing-surfacing  pattern.  This  appears  to  be  the  case  in 
adult-calf  associations.  Porpoise  #4  and  its  calf  (  #5)  and  #52  and  its  calf  (#50) 
often  surfaced  together.  In  the  #4-calf  association,  this  relationship  was  still  strong 
in  December,  1976,  approximately  two  and  one-half  years  after  #4's  calf  was  born. 
Nevertheless,  such  association  is  not  absolute.  As  the  increased  surfacing  rate  of 
#50  (17.8  sec/dive)  over  that  of  its  presumed  mother  (#52  —  21.3  sec/dive) 
showrs,  the  calf  at  times  surfaced  independently  of  the  adult.  When  it  did  so,  it  was 
almost  always  seen  moving  ahead  of  the  subgroup  of  animals,  an  apparently  in- 
vestigative or  "play"  behavior  summarized  for  other  species  by  Norris  and  Dohl 
(1978a).  While  Caldwell  and  Caldwell  (1972)  reported  the  same  type  of  non- 
association  to  be  present  at  times  in  captive  porpoises,  Irvine  (University  of 
Florida,  personal  communication)  believes  that  it  does  not  occur  in  a  bottlenose 
porpoise  population  in  the  Sarasota-Bradenton  area  of  central  west  Florida.  A 
possible  explanation  of  this  difference  in  behavior  may  be  the  relatively  undisturbed 
state  of  porpoises  in  the  present  study,  unlike  Florida  populations  which  have  been 
harassed  by  capture  vessels  and  tourist  boats.  Thus,  Florida  porpoises  may  keep 
their  young  within  the  confines  of  the  school  during  periods  of  possible  danger  such 
as  the  approach  of  a  boat,  while  no  such  restriction  appeared  to  apply  to  porpoises 
in  Golfo  San  Jose.  Instead,  calves  and  subadults  at  times  briefly  left  the  side  of 
the  adult  with  which  they  normally  surfaced  and  approached  the  investigator's  boat 
without  apparent  caution. 

Since  bottlenose  porpoises  have  an  approximately  12-month  gestation  period 
(Sergeant,  Caldwell,  and  Caldwell,  1973),  the  marked  summer  calving  season 
within  the  present  population  also  indicates  an  increase  in  mating  activity  at  that 
time.  Various  investigators  (Mead,  1975;  Oclell,  1975;  Sergeant,  Caldwell,  and 
Caldwell,  1973 ;  Evans,  Navy  Underwater  Center,  San  Diego,  personal  communica- 
tion, for  Dclphiniis  del  phis;  Nishiwaki,  Nakajima,  and  Kamiya,  1965,  and  Harrison, 
Brownell,  and  Boice,  1972,  for  Stcnclla  a/ tomato)  have  reported  a  tendency  toward 
bimodal  calving,  with  peaks  in  spring  and  fall.  Ridgway  and  Green  (1967)  found 
anatomical  evidence  for  mating  peaks  in  spring  and  fall  by  an  increase  in  testes 
weights  of  male  Delphinus  del  phis  and  Lagenorhynchus  ubliquidens  during  these 


PORPOISE  GROUP  ORGANIZATION  357 

two  seasons.  The  present  population  may  exhibit  a  similar  mating  and  calving 
trend,  with  one  calf  first  observed  in  November,  and  six  others  first  seen  from 
January  through  April.  Why  a  bimodal  calving  peak  appears  to  be  present  in 
various  different  species  of  toothed  cetaceans  is  not  known.  In  the  present  study, 
the  late  summer  calving  preference  coincides  with  the  highest  water  temperatures  of 
the  year.  This  higher  ambient  temperature,  as  in  most  terrestrial  mammals  and  in 
pinnipeds  and  baleen  whales,  may  be  of  physiological  advantage  to  the  newly-born 
young. 

The  present  study  demonstrates  that  by  systematically  photographing  small 
groupings  of  coastal  porpoises  much  can  be  learned  about  their  organization  and 
dynamics.  This  represents  one  of  the  first  times  that  such  an  attempt  has  been 
reported  (see  also  Wiirsig  and  Wiirsig,  1977),  and  it  is  hoped  that  more  such 
studies,  on  different  odontocete  cetacean  populations  and  on  different  species,  will 
soon  be  made.  In  this  manner,  by  observing  coastal  porpoises  for  long  periods  of 
time,  long-overdue  descriptions  of  natural  populations — analogous  to  the  recent 
flowering  of  primate  research — may  take  place. 


Dr.  Charles  Walcott  provided  encouragement  and  advice  throughout  the  study, 
and  Dr.  Roger  Payne  supported  all  phases  of  the  field  work.  Peter  Tyack,  Martin 
Hyatt,  Russ  Charif,  Christopher  Clark.  Jane  Moon,  Hugo  Callejas,  and  Carlos 
Garcia  assisted  with  the  gathering  of  field  data.  Mary  Griswold  Smith  of  the 
National  Geographic  Society  arranged  for  generous  help  with  35  mm  film,  and  gave 
advice  on  how  best  to  utilize  it.  Jan  Wolitzky  and  Steven  Ferraro  provided  in- 
valuable assistance  with  computer  analyses,  and  Dr.  Charles  Walcott  critically  read 
the  manuscript.  This  study  was  supported  by  contributions  and  facilities  from  the 
New  York  Zoological  Society,  by  the  Program  for  Neurobiology  and  Behavior  of 
the  State  University  of  New  York  at  Stony  Brook,  and  by  grants  to  R.  Payne  and 
C.  WTalcott  from  the  Committee  for  Research  and  Exploration  of  the  National 
Geographic  Society.  Melany  Wiirsig  assisted  in  all  phases  of  the  study. 

SUMMARY 

1.  During  a  21-month  study,  individuals  of  Tursiops  tnincalns  in  Golfo  San  Jose, 
Argentina,  exhibited  a  four  month  occurrence  cycle,  but  showed  no  seasonal  migra- 
tion. 

2.  Subgroups  numbering  S  to  22  animals  included  a  small  core  unit  of  indi- 
viduals which  were  consistently  found  together.     Other  animals  appeared  and  dis- 
appeared in  these  subgroups  on  different  days  in  a  highly   fluid  manner  which 
paralleled  the  open  society  of  African  chimpanzees.  Pan  troglodytes. 

3.  Some  adults  showed  weak  and  changing  surfacing  associations  with  other 
adults.    Calves  consistently  surfaced  together  with  a  particular  adult,  except  during 
apparent  play  or  investigative  behavior,  when  calves  left  adults  for  brief  periods. 
The  mean  dive  time  per  animal  was  21.8  seconds. 

4.  Six  of  seven  calves  were  born  in  late  summer.     This  calving  peak  coincided 
with  the  highest  water  temperatures  of  the  year. 


358  BERND  WURSIG 

LITERATURE  CITKI) 

CM. DWELL,  D.  K.,  AND  M.  C.  CALDWELL,   1972.     The  world  of  the  bottlenosed  dolphin.    J.   B. 

Lippincott  Press,  Philadelphia  and  New  York,  157  pp. 
DOUGLAS-HAMILTON,    I.,   AND   O.   DOUGLAS-HAMILTON,    1975.     Amain/    the   elephants.    Viking 

Press,  New  York,  284  pp. 
EVANS,  \V.  E.,  1971.     Orientation  behavior  of  delphinids:    radio  telemetric  studies.  Ann.  N.   Y. 

A  cad.  Sci.,  188:  142-160. 
EVANS,   W.   E.,   1974.     Radio-telemetric   studies   of  two  species   of   small  odontocete  cetaceans. 

Pages  386-394  in  W.  Schevill,  Ed.,  The  in'halc  problem;  <i  status  report.    Harvard  Uni- 
versity Press,  Cambridge,  Massachusetts. 
GOODALL,  J..  1965.     Chimpanzees  of  the  Gombe  Stream  Reserve.     Pages  425-473  in  I.  DeVore, 

Ed.,  Primate  behavior.    Holt,  Rinehart,  and  Winston,  New  York. 
GREENWOOD,  J.  A.,  AND  D.  DURAND,   1955.     The  distribution  of  length  and  components  of  the 

sum  of  n  random  unit  vectors.    Ann.  Matli.  Stat.,  26:  233-246. 
GUNTER,   G.,   1942.     Contributions  to  the  natural   history   of  the   bottlenose   dolphin,    Tursiops 

truncatiis  (Montague),  on  the  Texas  coast,  with  particular  reference  to  food  habits.  /. 

Mammal. ,23:  267-276. 
HALL,  K.  1968.     Social  organization  of  the  old-world  monkeys  and  apes.    Pages  7-31  in  P.  Jay, 

Ed.,   Primates:    studies  in  adaptation   and  -variability.    Holt,    Rinehart,   and   Winston, 

New  York. 
HARRISON,  R.  J.,  R.  L.  BROWNELL,  JR.,  AND  R.  C.  BOICE,   1972.     Reproduction  and  gonadal 

appearances  in  some  odontocetes.     Pages  361-429  in   R.  J.   Harrison,   Ed.,  Functional 

anatomy  of  marine  niammals,  Vol.  1.    Academic  Press,   New  York. 
MEAD,  J.  G.,   1975.     Preliminary  report  on  the  former  net   fisheries  for   Tursiops  fntncatiis  in 

the  western  north  Atlantic.   /.  Fish.  Res.  Board  Can.,  32:  1155-1162. 
MITCHELL,    E.,    1975.      Porpoise,    dolphin,    and    small    whale    fisheries    of    the    world.     Umvin 

Brothers,  Surrey,  England,  pages  30-101. 
NISHIDA,  T.  1968.     The  social  group  of  wild  chimpanzees  in  the  Mahali  mountains.    Primates. 

9:  175-198. 
NISHIWAKI,  M.,  M.  NAKAJIMA,  AND  T.  KAMIVA,   1965.     A  rare  species  of  dolphin   (Stenella 

attenuata)  from  Arari,  Japan.    Sci.  Rep.  ll'hales  Res.  lust.   (Tokyo),  19:  53-64. 
NORRIS,  K.  S.,  1974.     The  porpoise  watcher.    W.  W.  Norton,  New  York,  250  pp. 
NORRIS,  K.  S.,  AND  T.  P.  DOHL,  1978a.     The  structure  and  functions  of  cetacean  schools.     In 

press  in  L.  Herman,  Ed.,  Cetacean  behavior.    Wiley  Interscience,  New  York. 
NORRIS,   K.   S.,   AND   T.    P.    DOHL,    1978b.     The   behavior   of   the   Hawaiian    spinner   porpoise, 

Stenella  lonuirostris.  LI.  S.  Natl.  Mar.  Fish.  Serv.  Fish.  Bull.,  in  press. 
ODELL,  D.  K.,  1975.     Status  and  aspects  of  the  life  history  of  the  bottle  nose  dolphin,   Tursiops 

truncatiis,  in  Florida.    J.  Fish.  Res.  Board  Can..  32:   1055-1058. 
PERRIN,  W.  F.,  1975.     Distribution  and  differentiation  of  populations  of  dolphins  of  the  genus 

Stenella  in  the  eastern  tropical  Pacific.    /.  Fish.  Res.  Board  Can.,  32  :   1059-1067. 
REYNOLDS,  V.,  1965.     Some  behavioral  comparisons  between  the  chimpanzee  and  the  mountain 

gorilla  in  the  wild.    Am.  Anthropol.,  67  :  691-706. 
REYNOLDS,  V.,  AND  F.  REYNOLDS,  1965.     Chimpanzees  of  the  Budongo  Forest.     Pages  368-424 

in  I.  DeVore,  Ed.,  Primate  behavior.    Holt,  Rinehart,  and  Winston,  New  York. 
RIDGWAY,   S.  H.,  AND  R.  F.   GREEN,   1967.     Evidence  for  a  sexual   rhythm  in  male  porpoises, 

Lagenorhynchus  obliquidens  and  Delpliinus  delpliis  bairdi.  Nor.  Hvalfanc/st — tidcndc,  1: 

1-8. 
ROHLF,  F.  J.,  AND  R.  R.  SOKAL,  1969.     Statistical  tables.    W.  H.  Freeman,  San  Francisco,  253 

pp. 
SAAYMAN,  G.   S.,  D.   BOWKK.  AND  C.  K.  TAYLER,   1972.     Observations  on  inshore  and  pelagic 

dolphins  on  the  south-eastern  Cape  coast  of  South  Africa.    Kocdoc,  15:  1-24. 
SERGEANT,  D.  E.,  D.  K.  CALDWELL,  AND  M.  C.  CALDWELL,  1973. .  Age,  growth,  and  maturity  of 

bottlenosed    dolphins    (Tursiops    truncatus)    from    north-east    Florida.     /.    Fish.    Res. 

Board  Can.,  30:  1009-1011. 

SOKAL,  R.  R.,  AND  F.  J.  ROHLF,  1969.     Biometry.    W.  H.  Freeman,  San  Francisco,  776  pp. 
SUGIYAMA,    Y.,    1968.    Social    organization   of   chimpanzees    in    the    Budongo    Forest,    Uganda. 

Primates,  9 :  225-258. 


PORPOISE  GROUP  ORGANIZATION  359 

TAVOLGA,  M.  C,  1966.  Behavior  of  the  bottlenose  dolphin  (Tin-slops  tnuicatus)  :  social  inter- 
actions in  a  captive  colony.  Pages  718-730  in  K.  Norris,  Ed.,  ll'lmlcs.  dolphins,  and 
porpoises.  University  of  California  Press,  Los  Angeles. 

TAYLER,  C.  K.,  AND  G.  S.  SAAYMAN,  1972a.  A  method  for  determining  the  composition,  de- 
ployment, and  stability  of  free-ranging  dolphins.  Z.  Sacugeticrkd.,  37:  116-119. 

TAYLER,  C.  K.,  AND  G.  S.  SAAYMAN,  1972b.  The  social  organization  and  behaviour  of  dolphins 
(Tursiops  aduncus)  and  baboons  (Papio  ursinus)  :  some  comparisons  and  assessments. 
Ann.  Cape  Prov.  Mus.  Nat.  Hist.,  9:  11-49. 

TRUE,  F.  W.,  1891.  Observations  on  the  life  history  of  the  bottlenose  porpoise.  U.  S.  Xntl. 
.\fus.  Proc.,  1890:  197-203. 

WILSON,  E.  O.,  1975.  Sociobiology.  the  new  synfhcsis.  Harvard  University  Press,  Cambridge, 
Massachusetts,  697  pp. 

WURSIG,  B.,  AND  M.  WURSIG,  1977.  The  photographic  determination  of  group  size,  composition, 
and  stability  of  coastal  porpoises,  Tursiops  truncatus.  Science,  198:  755-756. 


Continued  from  Cover  Two 

of  biological  journal  titles  is  that  published  each  year  by  BIOLOGICAL  ABSTRACTS  (BIOSIS  List  of 
Serials;  most  recent  issue,  1976).  Foreign  authors,  and  others  who  are  accustomed  to  use 
THE  WORLD  LIST  OF  SCIENTIFIC  PERIODICALS,  may  find  a  booklet  published  by  the  Biological 
Council  of  the  U.K.  (obtainable  from  the  Institute  of  Biology,  41  Queen's  Gate,  London,  S.W.7, 
England,  U.K.  at  £0.65  or  $1.75)  useful,  since  it  sets  out  the  WORLD  LIST  abbreviations  for  most 
biological  journals  with  notes  of  the  USASI  abbreviations  where  these  differ.  CHEMICAL  AB- 
STRACTS publishes  quarterly  supplements  of  additional  abbreviations.  The  following  points  of 
reference  style  for  THE  BIOLOGICAL  BULLETIN  differ  from  USASI  (or  modified  WORLD  LIST) 
usage: 

A.  Journal  abbreviations,  and  book  titles,  all  underlined  (for  italics) 

B.  All  components  of  abbreviations  with  initial  capitals  (not  as  European  usage  in  WORLD 
LIST  e.g.  J.  Cell.  Comp.  Physiol.  NOT  /.  cell.  comp.  Physiol.) 

C.  All  abbreviated  components  must  be  followed  by   a  period,  whole  word  components 
must  not  (not  strictly  as  USASI  usage,  i.e.  J.  Cancer  Res.) 

D.  Space  between  all  components  (e.g.  J.  Cell.  Comp.  Physiol.  not  J. Cell. Comp. Physiol.) 

E.  We  strongly  recommend  that  more  unusual  words  in  journal  titles  be  spelled  out  in  full, 
rather  than  employing  lengthy,  peculiar  "abbreviations"  or  new  abbreviations  invented  by  the 
author.     For  example,  use  Rit  Visindafjelags  Islendinga  without  abbreviation.     Even  in  more 
familiar  languages,  Z.   Vererbungslehre  is  preferred  to  Z.   VerebLehre  (WORLD  LIST)  or  Z.   Verer- 
bungsl.   (USASI).     Accurate  and  complete  communication  of  the  reference  is  more  important  than 
minor  savings  in  printing  costs. 

F.  All  single  word  journal  titles  in  full  (e.g.  Veliger,  Ecology,  'Brain). 

G.  The  order  of  abbreviated  components  should  be  the  same  as  the  word  order  of  the  com- 
plete title  (i.e.  Proc.  and  Trans,  placed  where  they  appear,  not  transposed  as  in  some  BIOLOGICAL 
ABSTRACTS  listings). 

H.  Spell  out  London,  Tokyo,  Paris,  Edinburgh,  Lisbon,  etc.  where  part  of  journal  title. 
I.  Series  letters  etc.  immediately  before  volume  number. 

J.  A  few  well-known  international  journals  in  their  preferred  forms  rather  than  WORLD 
LIST  or  USASI  usage  (e.g.  Nature,  Science,  Evolution  NOT  Nature,  Land.;  Science,  N.  Y.;  Evolution, 
Lancaster,  Pa.) 

K.  The  correct  abbreviation  for  THE  BIOLOGICAL  BULLETIN  is  Biol.  Bull. 

5.  Figures.     The  dimensions  of  the  printed  page,  5  by  7f  inches,  should  be  kept  in  mind  in 
preparing  figures  for  publication.     Illustrations  should  be  large  enough  so  that  all  details  will  be 
clear  after  appropriate  reduction.     Explanatory  matter  should  be  included  separately  in  legends 
as  far  as  possible,  although  the  axes  should  always  be  numbered  and  identified  on  the  illustration 
itself.     Figures  should  be  prepared  for  reproduction  as  either  line-cuts  or  halftones;  no  other 
methods  will  be  used.     Figures  to  be  reproduced  as  line-cuts  should  be  drawn  in  black  ink  on  white 
paper,  good  quality  tracing  cloth  or  plastic,  or  blue-lined  coordinate  paper;  those  to  be  reproduced 
as  halftones  should  be  mounted  on  board,  and  both  designating  numbers  or  letters  and  scale-bars 
should  be  affixed  directly  on  the  figures.     We  recommend  that  halftones  submitted  to  us  be 
mounted  prints  made  at  about  1%  times  the  linear  dimensions  of  the  final  printing  desired  (the 
actual  best  reductions  are  achieved  from  copy  in  the  range  from  1^  to  2  times  the  linear  dimen- 
sions).    As  regards  line-blocks,  originals  can  be  designed  for  even  greater  reductions  but  are  best 
in  the  range  1?  to  3  times.     All  figures  should  be  numbered  in  consecutive  order,  with  no  distinc- 
tion between  text  and  plate-figures.     The  author's  name  should  appear  on  the  reverse  side  of  all 
figures,  and  the  inked  originals  for  line-blocks  must  be  submitted  for  block-making. 

6.  Mailing.     Manuscripts  should  be  packed  flat.     All  illustrations  larger  than  85  by  1 1  inches 
must  be  accompanied  by  photographic  reproductions  or  tracings  that  may  be  folded  to  page  size. 

Reprints.  Reprints  may  be  obtained  at  cost;  approximate  prices  will  be  furnished  by  the 
Managing  Editor  upon  request. 


CONTENTS 


BRADLEY,  BRIAN  P. 

Increase  in  range  of  temperature  tolerance  by  acclimation  in  the 
copepod  Eurytemora  affinis 177 

DAME,  R.  F.  AND  F.  J.  VERNBERG 

The  influence  of  constant  and  cyclic  acclimation  temperatures  on  the 
metabolic  rates  of  Panopeus  herbstii  and  Uca  pugilator 188 

DEL  PINO,  EUGENIA  M.,  AND  A.  A.  HUMPHRIES,  JR. 

Multiple  nuclei  during  early  oogenesis  in  Flectonotus  pygmaeus 

and  other  marsupial  frogs 198 

FISHER,  FRANK  M.,  JR.  AND  JOHN  A.  OAKS 

Evidence  for  a  nonintestinal  nutritional  mechanism  in  the  rhyn- 
chocoelan,  Linens  ruber 213 

FUZESSERY,  ZOLTAN  M.,  WILLIAM  E.  S.  CARR,  AND  BARRY  W.  ACHE 

Antennular  chemosensitivity  in  the  spiny  lobster,  Panulirus  argus: 
studies  of  taurine  sensitive  receptors 226 

GOY,  JOSEPH  W.  AND  ANTHONY  J.  PROVENZANO,  JR. 

Larval  development  of  the  rare  burrowing  mud  shrimp  Naushonia 
crangonoides  Kingsley  (Decapoda  :  Thalassinidea;  Laomediidae)  241 

HINES,  ANSON  H. 

Reproduction  in  three  species  of  intertidal  barnacles  from  central 
California ??-v*U 262 

PECHENIK,  JAN  A. 

Adaptations  to  intertidal  development :  studies  on  Nassarius 
obsoletus 282 

PRUSCH,  ROBERT  D.  AND  CAROL  HALL 

Diff usional  water  permeability  in  selected  marine  bivalves 292 

ROBERTSON,  DOUGLAS  R. 

The  light-dark  cycle  and  a  nonlinear  analysis  of  lunar  perturbations 
and  barometric  pressure  associated  with  the  annual  locomotor 
activity  of  the  frog,  Rana  pipiens 302 

SHIRLEY,  THOMAS  C.,  GUY  J.  DENOUX,  AND  WILLIAM  B.  STICKLE 

Seasonal  respiration  in  the  marsh  periwinkle,  Littorina  irrorata. .  . .     322 

STEPHENS,  GROVER  C.,  MARVA  J.  VOLK,  STEPHEN  H.  WRIGHT,  AND 
PETER  S.  BACKLUND 

Transepidermal  accumulation  of  naturally  occurring  amino  acids  in 

the  sand  dollar,  Dendr aster  excentricus 335 

WURSIG,  BERND 

Occurrence  and  group  organization  of  Atlantic  bottlenose  porpoises 
(Tursiops  truncatus)  in  an  Argentine  Bay 348 


Volume  154  Number  3 


THE 


BIOLOGICAL  BULLETIN 


PUBLISHED  BY 
THE  MARINE  BIOLOGICAL  LABORATORY 

j      Woods  H 


Editorial  Board 

EDWARD  M.  BERGER,  Dartmouth  College  MEREDITH  L.  JONES,  Smithsonian  Institution 

JOHN  M.  ANDERSON,  Cornell  University  HOWARD  A.  SCHNEIDERMAN,  University  of 

California,  Irvine 

JOHN  B.  BUCK,  National  Institutes  of  Health 

RALPH  I.  SMITH,  University  of  California, 

JOHN  D.  COSTLOW,  Duke  University 

F.  JOHN  VERNBERG,  University  of 
PHILIP  B.  DUNHAM,  Syracuse  University  South  Carolina 

J.  B.  JENNINGS,  University  of  Leeds  CARROLL  M.  WILLIAMS,  Harvard  University 

W.  D.  RUSSELL-HUNTER,  Syracuse  University 
Managing  Editor 


JUNE,  1978 


" 


Printed  and  Issued  by 
LANCASTER  PRESS,  Inc. 

PRINCE  &  LEMON  STS. 
LANCASTER,  PA. 


THE  BIOLOGICAL  BULLETIN 

THE  BIOLOGICAL  BULLETIN  is  published  six  times  a  year  by  the  Marine  Biological  Laboratory, 
MBL  Street,  Woods  Hole,  Massachusetts  02543. 

Subscriptions  and  similar  matter  should  be  addressed  to  THE  BIOLOGICAL  BULLETIN,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts.  Agent  for  Great  Britain:  Wheldon  and 
Wesley,  Limited,  2,  3  and  4  Arthur  Street,  New  Oxford  Street,  London,  W.  C.  2.  Single  numbers, 
$8.00.  Subscription  per  volume  (three  issues),  $22.00,  (this  is  $44.00  per  year  for  six  issues). 

Communications  relative  to  manuscripts  should  be  sent  to  Dr.  W.  D.  Russell-Hunter,  Marine 
Biological  Laboratory,  Woods  Hole,  Massachusetts  02543  between  June  1  and  September  1,  and 
to  Dr.  W.  D.  Russell-Hunter,  P.O.  Box  103,  University  Station,  Syracuse,  New  York  13210, 
during  the  remainder  of  the  year. 


Copyright   ©  1978,  by  the  Marine  Biological  Laboratory 
Second-class  postage  paid  at  Woods  Hole,  Mass.,  and  additional  mailing  offices. 


INSTRUCTIONS  TO  AUTHORS 

THE  BIOLOGICAL  BULLETIN  accepts  original  research  reports  of  intermediate  length  on  a  variety 
of  subjects  of  biological  interest.  In  general,  these  papers  are  either  of  particular  interest  to  workers 
at  the  Marine  Biological  Laboratory,  or  of  outstanding  general  significance  to  a  large  number  of 
biologists  throughout  the  world.  Normally,  review  papers  (except  those  written  at  the  specific 
invitation  of  the  Editorial  Board),  very  short  papers  (less  than  five  printed  pages),  preliminary 
notes,  and  papers  which  describe  only  a  new  technique  or  method  without  presenting  substantial 
quantities  of  data  resulting  from  the  use  of  the  new  method  cannot  be  accepted  for  publication.  A 
paper  will  usually  appear  within  four  months  of  the  date  of  its  acceptance. 

The  Editorial  Board  requests  that  manuscripts  conform  to  the  requirements  set  below; 
those  manuscripts  which  do  not  conform  will  be  returned  to  authors  for  correction  before  review 
by  the  board. 

1.  Manuscripts.     Manuscripts  must  be  typed  in  double  spacing  (including  figure  legends, 
foot-notes,  bibliography,  etc.)  on  one  side  of  16-  or  20-lb.  bond  paper,  8|  by  11  inches.     They 
should  be  carefully  proof-read  before  being  submitted  and  all  typographical  errors  corrected 
legibly  in  black  ink.     Pages  should  be  numbered.     A  left-hand  margin  of  at  least  1£  inches 

should  be  allowed. 
^ 

2.  Tables,  Foot-Notes,  Figure  Legends,  etc.     Tables  should  be  typed  on  separate  sheets  and 
placed  after  the  Literature  Cited.     Because  of  the  high  cost  of  setting  such  material  in  type 
authors  are  earnestly  requested  to  limit  tabular  material  as  much  as  possible.     Similarly,  foot- 
notes to  tables  should  be  avoided  wherever  possible.     If  they  are  essential,  they  should  be  indi- 
cated by  asterisks,  daggers,  etc.,  rather  than  by  numbers.     Foot-notes  are  not  normally  permitted 
in  the  body  of  the  text.     Such  material  should  be  incorporated  into  the  text  where  appropriate. 
Explanations  of  figures  should  be  typed  double-spaced  and  placed  on  separate  sheets  at  the  end 
of  the  paper. 

3.  A  condensed  title  or  running  head  of  no  more  than  35  letters  and  spaces  should  be  included. 

4.  Literature  Cited.     The  list  of  references  should  be  headed  LITERATURE  CITED, 
should  conform  in  punctuation  and  arrangement  to  the  style  of  recent  issues  of  THE  BIOLOGICAL 
BULLETIN,  and  must  be  typed  double-spaced  on  separate  pages.     Note  that  citations  should 
include  complete  titles  and  inclusive  pagination.     Journal  abbreviations  should  normally  follow 
those  of  the  U.  S.  A.  Standards  Institute  (USASI),  as  adopted  by  BIOLOGICAL  ABSTRACTS  and 
CHEMICAL  ABSTRACTS,  with  the  minor  differences  set  out  below.     The  most  generally  useful  list 

Continued  on  Cover  Three 


Vol.  154,  No.  3  June,  1978 

THE 

BIOLOGICAL  BULLETIN 

PUBLISHED   BY   THE    MARINE   BIOLOGICAL   LABORATORY 


RHEOTAXIS  AND  CHEMORECEPTION   IN  THE  FRESHWATER 

SNAIL  BIOMPHALARIA   GLABRATA    (SAY)  :  ESTIMATION 

OF  THE  MOLECULAR  WEIGHTS 

OF  ACTIVE  FACTORS 

Reference:  Biol  Bull.,  154:  361-373.      (June,   1978) 

J.  D.  BOUSFIELD 

Population  and  Development  Group,  School  of  Biological  Science,  University  of  Sussex, 
Palmer,  Brighton,  BN1   9QG,  Sussex-,  England,   U.  K. 

For  many  aquatic  organisms  chemoreception  plays  an  important,  often  decisive 
role  in  the  selection  and  location  of  diet  items  (for  reviews  see  Kohn,  1961  ;  Len- 
hoff  and  Lindstedt,  1974;  Bardach,  1975).  In  many  instances  it  has  proved 
possible,  using  a  suitable  behavioral  or  physiological  assay,  to  identify  some  of 
the  chemicals  which  elicit  these  responses  (see  for  example  Carr,  1967;  Carr 
and  Chancy,  1976;  Gurin  and  Carr,  1971;  Kara,  1976,  1977;  Lenhoff.  1968, 
1969;  Suzuki  and  Tucker,  1971;  Pawson,  1977).  Indeed,  from  a  study  of  the 
relationships  between  chemical  structure  and  biological  activity  Hara  (1976,  1977), 
working  with  the  rainbow  trout,  Salino  gairdncri,  and  Lenhoff  (1968,  1969), 
with  the  marine  hydra.  Hydra  littoralis,  have  been  able  to  describe  in  some  detail 
the  physico-chemical  properties  of  the  receptor  sites  themselves.  The  long  term 
object  of  the  present  investigation  has  been  to  arrive  at  a  similar  level  of  under- 
standing of  the  chemosensory  mechanisms  employed  by  Bivmphalaria  glabrata 
(Say)  (Planorbidae,  Mollusca),  a  phytophagous  freshwater  snail  found  in  South 
America,  and  parts  of  the  Caribbean. 

The  need  for  a  detailed  study  of  chemoreception  in  this  organism  is  of  particular 
importance,  for  two  reasons.  First,  while  freshwater  plants  are  known  to  be  an 
important  factor  conditioning  the  habitats  of  many  species  of  freshwater  snails 
(Gaevskaya,  1969;  Bovbjerg,  1965;  Pimentel  and  White,  1959;  Pip  and  Stewart, 
1976)  little  is  known  about  the  chemosensory  basis  of  such  interactions.  Biom- 
phalaria  glabrata  is  particularly  suitable  for  such  a  study  for  it  has  been  shown 
to  orient  both  chemotactically  (Etges,  1963a,  b;  Michelson,  1960;  Townsend, 
1973a,  b,  1974)  and  rheotactically  (Etges  and  Frick,  1966)  to  dilute  solutions 
of  various  plant  extracts.  Secondly,  many  species  of  freshwater  snails  are  of 
considerable  economic  and  medical  importance  as  intermediate  hosts  of  digenetic 
trematodes  parasitic  in  man  or  domestic  animals.  Bioinphalaria  glabrata,  for  ex- 
ample, is  host  to  the  human  schistosome,  Schistosoina  iinuisoni  (Sambon). 
Attempts  to  control  the  disease  usually  rely  on  the  use  of  molluscicides  to  remove 
or  reduce  snail  populations  in  areas  where  the  risk  of  transmission  is  particularly 

361 


362  J.  D.  BOUSFIELD 

high  (\Yebbe  and  Jordan,  1966;  Farooq,  1973).  There  are,  however,  numerous 
problems  associated  with  the  use  of  conventional  molluscicides,  among  these  being 
their  high  cost  (  I>erg,  1973;  Ritchie,  1973)  and  ecological  side  effects  (Shift 
and  (iarnett,  1961;  Ritchie,  1973).  A  detailed  knowledge  of  the  role  played 
by  chemoreception  in  determining  the  distribution  of  snails  within  habitats  may 
lead  to  more  efficient  and  acceptable  methods  of  control.  For  example,  if  baits 
could  be  formulated  which  could  lure  the  snails  to  sites  where  slow  release 
molluscicides  were  present,  a  saving  in  both  labor  and  cost  might  be  achieved 
(Etges,  1963a,  b;  Cardarelli,  1977). 

In   this  paper  a  preliminary   study  of  the  molecular  weight  characteristics  of 
chemicals  eliciting  rheotaxis  is  described. 


MATERIALS  AND  METHODS 


Methods 


Reproductively  mature  specimens  of  a  Venezuelan  albino  strain  of  Biompha- 
laria  glabrata,  weighing  400  ±  30  mg  (x±s.d),  were  selected  from  laboratory 
cultures  maintained  in  the  manner  described  by  Thomas  (1973).  These  indi- 
viduals were  then  kept  in  water  (see  below)  at  densities  of  20/liter  in  small 
plastic  buckets,  maintained  at  26°  C,  under  a  constant  (12L:  12D)  regime.  Each 
container  was  thoroughly  cleaned  out  each  day  and  fresh  water  (ionic  composition 
KHCCV- 0.037  HIM;  KNOS— 0.0495  HIM;  NaHCO3— 0.634  HIM;  MgSO4— 0.13 
HIM;  Cad;. — 2  HIM;  preaerated  to  pH  7.8-8.2)  and  food  (0.1  g  fresh  lettuce  per 
snail  per  day)  provided.  Each  snail  was  used  in  no  more  than  one  trial  per  day, 
and  was  conditioned  in  clean  water  for  one  hour  before  the  test. 


Test  apparatus  and  assay  procedure 


The  test  arena  consisted  of  a  shallow  perspex  trough  (24  X  24  X  3  cm)  into 
which  a  translucent  polythene  cylinder  (23  cm  diameter  X  3  cm  deep)  was  fitted. 
The  substrate  was  formed  by  a  glass  plate  which  could  be  removed,  and  thoroughly 
cleaned  before  each  trial.  Snails  placed  at  the  center  of  this  plate  were  thus 
free  to  move  within  a  cylindrically  symmetric  environment. 

To  create  the  flow,  two  jets  of  compressed  air  were  directed  in  opposite 
directions  around  the  edges  of  the  container  from  a  single  T-tube  held  close  to 
and  parallel  with  the  surface.  The  net  effect  was  to  draw  a  stream  of  water 
approximately  12-15  cm  wide  and  with  surface  speed  0.75-1.5  cm/sec  across 
the  center  from  the  opposite  side  of  the  arena.  Flow  was  returned  via  the  sides. 

Only  snails  which  were  actively  moving  at  the  bottom  of  their  buckets  at  the 
time  of  their  test  were  used  in  this  study.  This  was  done  in  order  to  standardize, 
as  far  as  was  possible,  the  initial  behavioral  state  of  the  tested  individuals.  Each 
snail  was  removed  gently  from  its  container  and  then  held  just  above  the  center 
of  the  glass  plate  floor  until  the  foot  extended  and  adhered  to  it.  Subsequent 
movement  of  the  snails  was  plotted  on  tracing  paper  on  the  screen  of  a  video 
monitor.  This  was  connected  to  a  camera  observing  the  movement  of  the  snail 
from  beneath  the  arena. 


RHEOTAXIS  IN  BldMl'lfALARLl  GLABRATA  363 

All  tests  were  conducted  under  conditions  of  diffuse-  illumination  from  above. 
In  order  to  counteract  any  residual  directional  bias  due  either  to  phototaxis 
(Sodeman,  1973;  Sodeman  and  Dowda,  1974)  or  starting  configuration  (shell 
orientation/flow  direction)  snails  were  always  started  with  the  shell  parallel  to 
a  fixed  axis,  while  the  direction  of  the  current  was  rotated,  periodically,  through  90° 

Stimuli 

Lettuce  was  obtained  from  local  market  gardens.  Care  was  taken  to  ensure 
that  no  plants  had  been  treated  with  pesticides  in  the  interval  two  weeks  before  use. 
Commercially  available  wheatgerm  (Jordan's  Natural  Wheatgerm,  Holme  Mills, 
Biggleswade,  Bedfordshire)  was  used.  All  plant  extracts  were  made  on  the  eve 
of  an  experiment  using  a  standard  technique. 

In  the  case  of  lettuce,  20  g  (wet  weight)  of  the  outer  leaves  were  homogenized 
for  one  minute  in  a  laboratory  blender  in  50  ml  of  distilled  water  chilled  to  50°  C. 
This  was  filtered  through  a  Whatman  No.  1  paper,  made  up  to  100  ml  with  more 
distilled  water  and  centrifuged  at  10,000  rpm  for  30  minutes.  The  supernatant  was 
then  filtered  through  Whatman  glass  fiber  papers  GF/C  and  GF/F  (nominal 
cut-off  0.7  jum)  and  stored  at  5°  C  until  the  following  day. 

Wheatgerm  extract  was  made  using  the  same  technique,  but  here  the  working 
strength  wras  1  g  (dry  weight) /1 00  ml,  and  the  first  filtering  stage  was  omitted. 

Ultrafiltrates  were  prepared  using  Amicon  Diaflo®  ultrafiltration  membranes 
UM-05,  UM-2,  UM-10,  PM-30,  and  XM-100A.  According  to  Amicon  Corpora- 
tion (1974)  these  membranes  retain  microsolutes  of  molecular  weights  (mol  wt) 
greater  than  500,  1000,  10,000,  30,000,  and  100,000,  respectively.  Although 
retention  is  actually  a  function  of  molecular  size,  configuration  and  charge,  a 
nominal  retention  characteristic  curve  can  be  drawn  for  each  membrane,  and 
these  are  shown  in  Figure  1. 

Ultrafiltration  was  carried  out  at  5°  C  under  nitrogen  (20-40  psi)  using  a 
100  ml  stirred  cell.  All  membranes  were  flushed  in  situ  with  200  ml  distilled 
water  before  use.  One  hundred  ml  of  extract  was  prepared  as  described  above 
and  passed  through  the  cell  overnight. 

Stimulus  solutions  were  made  up  immediately  before  use  as  standard  extract 
dilutions  of  750  ml  of  water  preheated  to  26°  C.  This  filled  the  arena  to  a 
depth  of  approximately  1.5  cm.  As  soon  as  the  flow  pattern  had  stabilized,  the 
test  was  begun.  Controls  were  run  in  water  to  which  no  extract  had  been  added. 
In  order  to  prevent  contamination,  stimuli  were  used  for  only  one  test  and  then 
discarded,  the  apparatus  being  thoroughly  rinsed  between  tests  to  remove  any 
residual  traces  of  chemical. 

Statistics 

All  trails  representative  of  a  particular  treatment  were  superimposed  on  tracing 
paper  so  that  their  starting  positions  coincided,  and  the  flow  directions  were  parallel. 
The  individual  trails  were  then  scored  in  the  following  manner. 

The  overall  trial  direction  was  defined  as  the  angle  made  between  the  Hi  in- 
direction and  a  vector  joining  the  starting  point  and  the  place  where  the  trial 
first  cut  an  8  cm  radius  boundary  centered  on  the  middle  of  the  arena.  These 


364 


J.  D.  BOUSFIELD 


(V 


100 
80 
60 
40 
20 


0  L 


10' 


10' 


molecular 


weight 


FIGURE  1.  Molecular  weight  transmission  charateristics  of  five  Amico  Diaflo®  ultra- 
filtration  membranes  used  in  this  paper.  Data  plotted  in  semi-logarithmic  form.  The 
ordinate  shows  the  percentage  of  solute  of  given  molecular  weight  transmitted  by  the  mem- 
brane. Levels  of  the  transmission  (95%  +  5%)  are  indicated  by  dotted  lines.  [Re  plotted 
from  Atnicon  Corporation  publication  (1974,  see  Literature  Cited).] 

angles  were  measured  to  50°  accuracy  using  an  anticlockwise  convention  and  the 
flow  direction  as  reference. 

Test  control  data  distributions  were  compared  using  the  nonparametric  tests 
for  directional  data  developed  and  described  by  Watson  (1962)  and  Mardia 
(1972).  Overall  estimates  of  the  responses  were  obtained  by  treating  each 
datum  as  a  unit  vector  and  calculating  the  resultant  r  (r,  (9),  where 

data  set  =  (0;)  ;     i  =  1  to  n 
r  =  [(Lsin0i)2  + 

i  =  1 


6  =  tan-1 


sn 


cos  0i] 


However,  for  the  purpose  of  constructing  graphs  the  scalar  quantity  r  cos  6  is 
more  useful.  This  has  a  range  of  values  from  +1  to  --1  and  by  convention  has 
been  taken  as  positive  when  the  net  movement  is  upstream  (positive  rheotaxis) 
and  negative  when  the  net  movement  is  downstream  (negative  rheotaxis). 

RESULTS 

Typical  data  from  a  series  of  experiments  in  which  lettuce  extract    (cultivar 
Rcnatc)  was  tested  are  shown  in  Figure  2.     The  data  are  shown  in  two  forms. 


RHEOTAXIS  IN  BIOMPHALARIA  CLABRA'l.l  365 

To  the  top,  the  superimposed  trails  obtained  for  a  given  treatment  are  shown. 
Each  trail  is  from  a  different  snail,  and  in  each  case  the  movement  was  centrifugal, 
with  current  direction  from  6  o'clock  to  12  o'clock.  It  can  be  seen  that  lettuce 
extract  tested  at  a  concentration  of  10  ml/liter  produces  a  strong  polarization 
in  favor  of  upstream  movement  (positive  rheotaxis),  whereas  the  control  snails 
exhibited  very  little  directional  bias.  The  difference  between  these  two  distribu- 
tions was  highly  significant  [P  <  0.001,  Watson's  UL>  test,  (Watson,  1962)]. 
Below  these  trails  the  same  data  are  re-represented  in  the  form  of  a  circular 
histogram. 

Figure  2  also  shows  what  effect  passing  the  extract  through  ultrafiltration 
membranes  has  on  the  activity  of  the  solution.  As  can  be  seen  the  effect  of 
membranes  PM-30,  UM-10,  UM-2  and  UM-05  was  to  produce  a  gradual 
reduction  in  the  length  of  the  resultant  vector  and  an  increase  in  scatter  in  the 
individual  trail  directions  as  the  retention  characteristic  moved  to  progressively 
lower  molecular  weights.  At  this  concentration  (10  ml  extract/liter)  all  the 
treatments  produced  positive  responses  which  were  significantly  different  from 
the  controls  [PM-30,  UM-10,  P  <  0.01  ;  UM-2,  UM-05,  P  <  0.05 ;  Watson's 
U-  test  (Watson,  1962)].  The  activity  of  the  UM-05  and  UM-10  filtrates, 
however,  was  significantly  less  (P  <  0.01  and  0.05,  respectively)  than  that  of  the 
original  extract. 

The  reason  why  membranes  with  such  widely  different  characteristics  merely 
produce  a  gradual,  rather  than  an  all-or-nothing  effect  on  activity  can  be  seen 
immediately  from  Figure  1.  It  is  clear  that  there  is  a  considerable  overlap 
between  the  characteristics  of  the  four  membranes  concerned.  For  any  individual 
membrane  the  5—95%  transmission  limits  span  a  molecular  weight  range  equiva- 
lent to  at  least  one  order  of  magnitude.  Consequently,  in  order  to  be  able  to 
interpret  the  results  of  ultrafiltration,  some  method  of  calculating  the  attenuation 
produced  by  any  given  membrane  must  be  found. 

In  Figure  3  the  response  magnitude-extract  concentration  profile  for  Rcnatc 
lettuce  has  been  plotted  in  semi-logarithmic  form.  Other  forms  of  representation 
are  possible,  but  this  method  was  found  to  be  the  most  successful  for  the  pur- 
poses of  linearizing  the  data  (see  Beidler,  1971).  From  the  linear  regression 
on  these  points,  it  is  possible  to  calculate  the  concentration  of  lettuce  extract 
which  would  produce  a  response  equivalent  in  magnitude  to  the  response  pro- 
duced by  a  given  filtrate.  For  example,  UM-10  filtrate  tested  at  concentrations 
of  2  ml/liter  produces  a  response  which  is  equivalent  in  magnitude  to  that  of 
lettuce  extract  tested  at  a  concentration  of  only  0.4  ml/liter.  Thus  the  UM-10 
filtrate  only  contains  2Q%  of  the  original  activity.  From  Figure  1  it  can  be 
seen  that  the  point  of  20r/f  transmission  occurs  for  molecular  weights  of  approxi- 
mately 2000.  The  results  of  testing  four  different  ultrafiltrates  of  Rcnatc 
extract,  each  at  two  different  concentrations,  are  shown  in  Figure  3  (see  legend), 
and  the  molecular  weight  estimates  obtained  shown  in  Table  IA.  The  range  of 
this  extract  was  1000-10,000.  However,  since  it  has  been  shown  elsewhere 
(Carr,  Hall  and  Gurin,  1974)  that  stimulants  from  different  sources  may  be 
characterized  by  different  molecular  weight  spectrums,  these  results  have  been 
complemented  with  tests  using  a  different  cultivated  variety  of  lettuce  and  with 
wheatgerm.  Data  from  experiments  involving  lettuce  (cultivar  Amanda}  are 


366 


J.  D.  BOUSFIELD 


extract 


PM-30 


UM-10 


n=15 

r=0.82 

9=168° 


UM-2 


UM-05 


n=16 
r  =  0.48 
9=175° 


control 


FIGURE  2.  Trail  data  obtained  using  lettuce  extract  (cultivar  Rotate)  and  the  effect  pro- 
duced by  passing  it  through  Diaflo®  membranes.  All  stimuli  were  tested  at  concentrations 
of  10  ml  extract  or  ultrafiltrate  per  liter.  Control  values  were  obtained  using  water  alone. 
Each  data  set  is  shown  in  two  forms.  At  the  top  the  trails,  starting  at  the  center  and 
radiating  outward,  of  all  the  snails  (n)  tested  in  a  given  solution  are  shown  superimposed. 
Below  the  same,  data  are  represented  in  the  form  of  a  circular  histogram.  The  vector  in  the 
center  (r,  0)  represents  the  magnitude  of  the  resultant  of  the  individual  unit  vectors  obtained 
for  each  trial.  A  vector  reaching  the  edge  of  the  circle  signifies  a  case  in  which  all  the  snails 


KHEOTAXIS  IN  BIOMPHALARIA  GLABRATA 


367 


'1.0 


0> 

</> 
o 
u 


.0.5 


c 
o 

Q. 
V) 
(V 


-0.3  L 


0.1  1.0  10    ml/L 

log  (extract    concentration) 

FIGURE  3.  Response-concentration  profile  (circles)  obtained  for  lettuce  extract  (cultivar 
Rcnatc),  together  with  the  effects  produced  by  various  ultrafiltration  treatments  (other  sym- 
bols). The  projection  of  the  resultant  vector  on  the  direction  of  flow  (r  cos  ft]  is  used  as 
an  index  of  the  response  magnitude  ( see  Methods ) .  The  regression  equation  for  the  lettuce 
extract  data  is  r  cos  0  =  0.306  log  (concentration)  +0.48.  Each  extract  data  point  repre- 
sents the  resultant  of  35-45  individual  trials.  Ultrafiltrate  and  control  data  points  are  averages 
of  15-25  trials.  Squares  represent  PM-30  filtrate;  diamonds,  UM-10  filtrate;  triangles,  UM-2 
filtrates;  and  inverted  triangles,  UM-50  filtrate.  All  filtrates  were  tested  at  concentrations  of 
2  ml  and  10  ml/liter.  Control  values  are  shown  on  the  ordinate. 

also  shown  in  Table  IB.  Exactly  the  same  procedure  was  used  here.  The 
estimates  arrived  at  using  this  extract  agree  well  with  those  obtained  using  the 
Rcnatc  variety  (Table  IA).  The  range  of  values  obtained  was  from  1000  to  6000. 
In  contrast,  however,  estimates  for  the  molecular  weights  of  the  attractants 


moved  in  the  same  direction.  On  the  other  hand,  a  vector  of  near  zero  length  represents 
a  case  in  which  there  was  no  bias  and  the  direction  of  movement  tended  to  be  random.  In 
each  case  the  direction  of  flow  is  from  6  o'clock  to  12  o'clock.  Note  the  reduction  in  flow 
vector  length  and  in  the  clustering  of  the  trails  which  occurs  as  increasingly  more  retentive 
membranes  are  used. 


368 


J.  D.  BOUSFIELD 


TAHLE  I 

The  responses  obtained  for  various  ullrafiltrates  of  lettuce  and  wheatgerm  extracts,  together  with  estimates 
of  the  molecular  weights  of  the  attractants.  The  method  of  deriving  the  "estimated  concentration"  and 
"percentage  of  activity  left"  are  explained  in  the  text.  Where  the  response  to  the  filtrate  was  either  very 
similar  to  that  of  the  original  extract,  or  was  very  small  these  values  were  not  calculated,  but  a  minimum 
or  maximum  molecular  weight  estimate  is  given.  Asterisks  denote  values  obtained  using  retentates. 
Note  how  these  compare  with  those  obtained  using  filtrates. 


Stimulus 

Concen- 
tration 
(ml/liter) 

Response 
(r.  cos  6) 

Estimated 
cone, 
(ml  /liter) 

Per  cent 
activity 
left 

Estimated 
mol  wt 

(A)  Lettuce  (cultivar  Renate) 

UM-05  filtrate 

2 

-0.10 

— 

— 

>2000 

UM-05  filtrate 

10 

0.35 

0.40 

4 

1000 

UM-2  filtrate 

2 

0.09 

— 

— 

>3000 

UM-2  filtrate 

10 

0.48 

1.1 

11 

2000 

UM-10  filtrate 

2 

0.35 

0.40 

20 

2000 

UM-10  filtrate 

10 

0.48 

1.1 

11 

3000 

PM-30  filtrate 

2 

0.46 

0.80 

40 

10,000 

PM-30  filtrate 

10 

0.66 

4.0 

40 

10,000 

(B)  Lettuce  (cultivar  Amanda) 

UM-2  filtrate 

3 

-0.04 

— 

— 

>3000 

UM-2  filtrate 

8 

0.26 

0.56 

7 

2000 

UM-2  retentate 

8 

0.56 

8.0 

100 

>3000* 

UM-10  filtrate 

3 

0.11 

0.16 

5 

6000 

UM-10  filtrate 

8 

0.26 

0.56 

7 

4000 

UM-10  retentate 

8 

0.48 

4.0 

50 

1000* 

(C)  Wheatgerm 

UM-2  filtrate 

8 

-0.12 

— 

— 

>3000 

UM-10  filtrate 

8 

0.29 

0.1 

1.25 

>  10,000 

PM-30  filtrate 

2 

0.47 

0.4 

20 

30,000 

PM-30  filtrate 

8 

0.56 

0.8 

10 

40,000 

XM-100  filtrate 

2 

0.80 

— 

~1(M) 

~10,000 

XM-100  filtrate 

8 

0.77 

4 

50 

40,000 

in   wheatgerm   were,   on   the  whole,   higher    (mol   wt    10,000-40,000)    than   those 
obtained  using  lettuce  extract   (Table  1C). 


DISCUSSION 

Before  proceeding  with  a  discussion  of  the  results,  it  is  important  to  view 
critically  the  methods  used  in  this  study.  Ultrafiltration  has  been  used  extensively 
in  studies  of  chemoreception  as  a  means  of  removing  or  selectively  attenuating 
specific  molecular  weight  fractions  from  solutions  containing  stimulants  (Carr, 
Hall  and  Gurin,  1974;  Carr  and  Gurin,  1975;  Carr,  1976;  Carr  and  Chancy, 


RHEOTAXIS  IN  BIOMPHALARIA  GLABRAT.t  369 

1976).  However,  the  "cut-off  points"  for  each  ultrafiltration  membrane,  in  fact, 
span  a  considerable  range  of  molecular  weights  (Fig.  1).  Thus,  even  if  the 
stimulant  chemicals  were  all  of  the  same  molecular  weight,  a  membrane  whose 
"cut-off  point"  lay  above  that  weight  would  not  necessarily  remove  the  biological 
activity  completely.  The  situation  is  further  complicated  by  the  possibility  that 
the  "stimulant"  may,  in  fact,  be  a  group  of  chemicals  all  with  different  molecular 
weights. 

While  bearing  in  mind  that  the  published  characteristics  of  each  membrane 
are  at  best  nominal  and  depend  on  a  number  of  factors  such  as  molecular  charge, 
configuration  and  the  presence  of  other  solutes,  the  first  difficulty  may  be  over- 
come by  relating  the  activity  of  an  ultrafiltrate  to  the  stimulus  concentration-re- 
sponse magnitude  profile  of  the  original  extract.  This  provides  an  estimate  of 
the  alteration  produced  by  the  given  membrane  and  consequently  a  value  for  the 
molecular  weight  of  the  attractant.  The  second  difficulty,  the  possible  presence 
of  a  range  of  active  molecules  with  different  retention  characteristics,  can  be  over- 
come, to  some  extent,  by  using  a  range  of  membranes  with  widely  different  ultra- 
filtration  properties.  Membranes  whose  50%  retention  points  lies  above  the  mean 
molecular  weight  of  the  attractants  will  provide  estimates  of  this  mean  which 
are  too  high.  Conversely,  membranes  whose  5Q%  retention  points  lie  below 
this  mean  will  produce  estimates  which  are  too  low.  In  general  this  tendency 
for  ultrafiltration  membranes  with  high  molecular  weight  cut-offs  to  give  higher 
estimates  than  those  with  low  cut-offs  is  borne  out  by  the  results  shown  in 
Table  I,  although  the  effect  is  particularly  obvious  only  for  the  wheatgerm  data. 

The  results  of  this  preliminary  study  of  the  characteristics  of  stimulants 
triggering  rheotaxis  clearly  demonstrate  that  the  factors  involved  are  not  simple 
compounds,  such  as  amino  acids,  short  chain  organic  acids  or  small  sugars,  but 
are  substances  having  molecular  weights  in  excess  of  1000.  An  exact  value  is, 
however,  for  reasons  given  above,  difficult  to  determine.  While  the  estimates 
obtained  for  two  varieties  of  cultivated  lettuce  are  in  agreement  and  provide  a 
value  somewhat  below  10,000,  all  the  estimates  made  for  wheatgerm  lie  on  or 
above  this  limit  (Table  I).  It  is  unlikely  then  that  the  response  is  specific 
to  a  single  chemical  compound  as  has  been  found  to  be  the  case  in  some 
marine  coelenterates  (Lenhoff  and  Lindstedt,  1974).  On  the  contrary,  these 
differences  suggest  that  some  generalized  property  of  a  class  of  macromolecules 
is  the  active  stimulus.  It  is  interesting  to  note  in  this  context  that  differences 
in  stimulant  molecular  weights  have  also  been  found  for  the  shrimp  Palaemonetes 
pugio  when  tested  with  extracts  made  from  a  variety  of  marine  invertebrates 
(Carr  and  Gurin,  1975). 

In  the  past  studies  of  chemoreception  and  food-finding  behavior  in  aquatic 
organisms  have  stressed  the  role  played  by  low  molecular  weight  nitrogenous 
compounds.  For  example,  sensitivities  to  amines  and  amino  acids  have  been 
demonstrated  in  a  number  of  marine  and  freshwater  fish  (for  example  Carr, 
1976;  Carr  and  Chancy,  1976;  Kara,  1976,  1977;  Pawson,  1977;  Suzuki  and 
Tucker,  1971).  in  marine  Crustacea  (Fuzessery  and  Childress,  1975;  Laverack, 
1963;  Mackie,  1973),  marine  molluscs  (Carr,  1967;  Crisp,  1967;  Jahan-Parwar, 
1975)  and  a  freshwater  planarian  (Coward  and  Johannes,  1969).  Although  in 
many  instances  the  activity  of  food  extracts  is  well  accounted  for  by  the  presence 


.•570  J.  D.  BOUSFIELD 

of  these  substances  (sec  for  example  Carr,  1967,  1976;  Carr  and  Chancy,  1976; 
Mackie,  1973;  Pawson,  1977),  it  is  becoming  increasingly  clear  that  compounds 
of  larger  molecular  weights  play  an  important  stimulatory  role  in  certain  cases 
(see,  for  example,  Ash,  McClure  and  Hirsch,  1973;  Carr,  Hall  and  fiurin, 
1974;  Carr  and  ( iurin,  1975;  ( inrin  and  Carr,  1971).  For  instance,  Carr  and  his 
co-workers  have  shown  that  for  the  marine  prosobranch,  Nassarius  obsolctus, 
macromolecules  with  properties  consistent  with  those  of  proteins  and  peptides 
are  the  main  active  factors  in  extracts  eliciting  exploratory  feeding  behavior. 
In  the  fresh  water  planarian,  Dugesia  dorotoccphala,  the  factors  which  elicit  feed- 
ing behavior  have  molecular  weights  of  between  700  and  2000  (Ash  ct  a!.,  1973). 

Aquatic  macrophytes  and  algae  (Fogg,  1971;  Hellebust,  1974;  Wetzel  and 
Manny,  1972)  release  large  quantities  of  organic  carbon  into  the  surrounding 
water.  It  has  been  suggested  that  these  chemicals  may  attract  aquatic  snails  and 
be,  in  part,  responsible  for  certain  plant-snail  associations  observed  in  the  field 
(Pip  and  Stewart,  1976).  Natural  plant  exudates  may  also  be  responsible  for 
the  positive  rheotactic  movements  which  have  sometimes  been  observed  in  field 
mark-recapture  experiments  performed  with  B.  ylabrata  (Paulini,  1963;  Pimentel 
and  Ildefonson,  1957;  Radke  and  Ritchie,  1961).  They  are  certainly  not  simply 
responses  to  the  presence  of  the  currents  themselves  (Etges  and  Frick,  1966). 

The  majority  of  the  material  secreted  by  plants  is  made  up  of  low  molecular 
weight  compounds  such  as  glucose  and  glycollic  acid  (Hellebust,  1974;  Wetzel  and 
Manny,  1972)  but  polysaccharides,  polypeptides  and  glycoproteins  are  also  released 
(Fogg,  1971;  Hellebust,  1974).  These  simpler  compounds  are,  however,  often 
photolabile  and  may  be  rapidly  utilized  by  epiphytic  organisms  (Allen,  1976; 
Sepers,  1977;  Wetzel  and  Manny,  1972).  Since  the  rheotaxic  response  described 
here  allows  Biomphalaria  glabrata  to  orient  to  distant  sources  of  organic  chemical, 
it  is  possible  that  such  ecological  pressures  have  favored  the  evolution  of  chemo- 
receptivity  for  larger,  more  stable  molecules. 


This  work  was  supported  by  the  Scientific  Research  Council.  I  should  also 
like  to  thank  Drs.  P.  Benjamin,  P.  Harvey,  F.  McCapra,  and  J.  D.  Thomas  for 
their  valuable  advice  and  criticisms  during  both  the  experimental  work  and  prepara- 
tion of  the  manuscript. 

SUMMARY 

1.  Dilute  solutions  of  lettuce  and  wheatgerm  extracts  trigger  positive  rheotaxis 
in  the  freshwater  snail,  Biomphalaria  glabrata.     This  response  can  be  used  as  the 
basis  of  a  sensitive  bioassay  for  characterizing  and   identifying  the  chemicals  to 
which  the  snail  is  attracted. 

2.  Using  ultrafiltration  techniques  a  range  of  different  molecular  weight  frac- 
tions could  be  attenuated  or  removed  from  these  extracts.     By  comparing  the  activ- 
ity of  these  solutions  with  that  of  the  original  extract  an  estimate  of  the  molecular 
weight  of  the  attractant  could  be  made. 

3.  In  both  cases  the  molecular  weights  of  the  attractants  were  estimated   as 
being  greater  than  1000.    Those  in  the  lettuce  were  estimated  at  lying  between  1000 


RHEOTAX1S   IX  BIOMPHALARIA  GLABRATA  371 

and  10,000;  whereas  for  wheatgerm  the  values  were  slightly  higher  and  lay  between 
10,000  and  40,000.     The  ecological  significance  of  these  results  is  discussed. 

LITERATURE  CITED 

ALLEN,  H.  L.,  1976.  Dissolved  organic  matter  in  lakewater :  characteristics  of  molecular 
weight  size  factions  and  ecological  implications.  Oikos.  27:  64-70. 

AMICON  CORPORATION,  1974.  Selection  guide  and  catalogue.  Publication  No.  426,  Amicon  Ltd, 
57  Queens  Road,  High  Wycombe,  Bucks,  England. 

ASH,  J.  F.,  W.  O.  McCLVRE,  AND  J.  F.  HIRSCH,  1973.  Chemical  studies  of  a  factor  which 
elicits  feeding  behavior  in  Duyesia  dorotocephala.  Auiin.  Beh/n'..  21  :  796-800. 

BARDACH,  J.  E.,  1975.  Chemoreception  in  aquatic  animals.  Pages  121-132  in  D.  A.  Denton 
and  J.  P.  Coghlan,  Eds.,  Olfaction  and  taste — I-7.  Academic  Press,  London  and  New 
York. 

BEIDLER,  L.  M.,  1971.  Taste  receptor  stimulation  with  salts  and  acids.  Pages  200-220  in 
Handbook  of  sensory  physiology,  Volume  IV,  Chemical  senses.  Part  2,  I <iste. 
Springer-Verlag,  Berlin. 

BERG,  C.  O.,  1973.  Biological  control  of  snail  borne  diseases — a  review.  E.rp.  Parasito!.,  33  : 
318-330. 

BOVBJERG,  R.  V.,  1965.  Feeding  and  dispersal  in  the  snail  Stagnicola  rejie.ra  ( Basommato- 
phora :  Lymnaeidae).  Malacologia,  2:  199-207. 

CARDARELLI,  N.  F.,  1977.  Controlled  release  niolluscicides.  Environmental  Management  Labora- 
tory Monograph,  University  of  Akron,  Akron,  Ohio,  U.S.A. 

CARR,  W.  E.  S.,  1967.  Chemoreception  in  the  mud  snail,  Nassarius  obsoletus  II.  Identification 
of  stimulatory  substances.  Biol.  Bull.,  133  :  106-127. 

CARR,  W.  E.  S.,  1976.  Chemoreception  and  feeding  behavior  in  the  pigfish  Orthopristis 
clirysopterus :  characterisation  and  identification  of  stimulatory  substances  in  shrimp 
extract.  Coinp.  Biocliem.  Physiol.,  55A  :  153-157. 

CARR,  W.  E.  S.,  AND  T.  B.  CHANEY,  1976.  Chemical  stimulation  of  feeding  behavior  in  the 
pinfish  Laijodon  rhomboid  es:  identification  and  characterisation  of  stimulatory  sub- 
stances extracted  from  shrimp.  Coinp.  Biocliem.  Physiol.,  54A :  437-441. 

CARR,  W.  E.  S.,  AND  S.  GURIN,  1975.  Chemoreception  in  the  shrimp  Palaeinonetes  pugio: 
comparative  study  of  stimulatory  substances  in  human  serum.  Biol.  Bull.,  148:  380-392. 

CARR,  W.  E.  S.,  E.  R.  HALL,  AND  S.  GURIN,  1974.  Chemoreception  and  the  role  of  proteins: 
a  comparative  study.  Comp.  Biochcm.  Physiol.,  47A  :  559-566. 

COWARD,  S.  J.,  AND  R.  E.  JOHANNES,  1969.  Amino  acid  Chemoreception  by  the  planarian 
Dugesia  dorotocephala.  Comp.  Biocliem.  Physiol..  29  :  475-478. 

CRISP,  D.  J.,   1967.     Chemoreception  in  Cirripedes.     Biol.  Bull..  133  :   128-140. 

ETGES,  F.  J.,  1963a.  Experimental  studies  on  Chemoreception  and  klinokinetic  responses  of 
Australorbis,  Bulinus,  and  Helisoma  to  chemical  stimulation.  Am.  J.  Trap.  Med.  H\g,. 
12:  686-695. 

ETGES,  F.  J.,  1963b.  Effects  of  Schistosoma  mansoni  infection  on  chemosensitivity  and  orienta- 
tion of  Australorbis  t/labratus.  Am.  J.  Trop.  Med.  Hyg.,  12:  696-700. 

ETGES,  F.  J.,  AND  L.  P.  FRICK,  1966.  An  experimental  field  study  of  Chemoreception  and  re- 
sponse in  Australorbis  glabratus  (Say)  under  rheotactic  conditions.  Am.  J.  Trop. 
Med.  Hyg.,  15:  434-438.' 

FAROOQ,  M.,  1973.  Planning  and  organisation  of  control  programmes.  Pages  438-457  in 
N.  Ansari,  Ed.,  Epideiniolo(i\  and  control  of  Schistosomiasis  (Bilharsiasis) .  World 
Health  Organisation  (ISBN."  3-8055-1340-2),  S.  Karger,  Basel. 

FOGG,  G.  E.,  1971.  Extracellular  products  of  algae  in  freshwater.  Arch.  Hydrobiol.  Organ. 
Intermit.  I'erem.  Theor.  Angewandte  Limnol.,  5:  1-25. 

FUZESSERY,  Z.  M.,  AND  J.  J.  CuiLDRESS,  1975.  Comparative  chemosensitivity  to  amino  acids 
and  their  role  in  the  feeding  activity  of  bathypelagic  and  littoral  crustaceans.  Biol. 
Bull..  149:  522-538. 

GAEVSKAYA,  N.  S.,  1969.  The  role  of  higher  aquatic  plants  in  the  nutrition  of  the  animals 
of  freshwater  basins,  National  Lending  Library  for  Science  and  Technology,  Boston 
Spa,  England,  354  pp. 


372  J.  I).  BOUSFIELD 

GURIN,  A.,  AND  W.   E.   S.   CAKR,   1971.     Chemoreception   in  Nassarius  obsoletus:   the   role  of 

specific  stimulatory  proteins.     Science.  1974:  293-295. 
HAKA,   T.   J.,    1976.      Structure-activity   relationships   of  amino-acids   in   fish   olfaction.      Comp. 

Biocliein.  Physio!.,  54A  :  31-36. 
HAKA,  T.  J.,   1977.     Further  studies  on  the   structure-activity  relationships  of  amino  acids   in 

fish  olfaction.    Comp.  Biochcm.  Physiol..  56A  :  559-565. 
HELLEBUST,  J.  A.,   1974.     Extracellular  products.     Pages  838-863  in  W.   D.   P.   Steward,   Ed., 

Botanical  monographs..  Vol.  10,  Alt/ul  physiology  and  biochemistry.     Blackwell  Scien- 
tific Pub.    (IBSN.  0-632-09100-2),  Oxford. 

KOIIN,  A.  J.,   1961.     Chemoreception  in  gastropod  molluscs.     Am.  Zoo/.,   1  :   291-308. 
JAHAN-PARWAR,  B.,  1975.    Chemoreception  in  gastropods.     Pages  133-139  in  D.  A.  Denton  and 

J.  P.  Coghlan,  Eds.,  Olfaction  and  taste — V.    Academic  Press,  New  York  and  London. 
LAVERACK,  M.   S.,   1963.     Aspects  of  Chemoreception  in  Crustacea.     Comp.  Biochcm.  Physiol., 

8:   141-151. 

LENHOFF,  H.  M.,  1968.     Behavior,  hormones  and  Hydra.     Science,  161  :  434-442. 
LENHOFF,   H.    M.,    1969.      pH    profile   of   a   peptide    receptor.      Comp.    Biochcm,   Physiol..   28: 

571-586. 
LENHOFF,  H.  M.,  AND  K.  J.  LINDSTEDT,   1974.     Chemoreception  in  aquatic  invertebrates   with 

special  emphasis  on  the  feeding  behavior  of  coelenterates.      Pages    143-175   in   P.   T. 

Grant    and    A.    M.    Mackie,    Eds.,    Chemoreception    in    marine    organisms.      Academic 

Press,  New  York  and  London. 
MACKIE,  A.  M.,  1973.     The  chemical  basis  of  food  detection  in  the  lobster  Homarus  gammarus. 

Mar.  Biol.,  21  :  103-108. 
MARDIA,  K.  V.,  1972.     Statistics  of  directional  data.    Academic  Press,   (ISBN.  0-12-471150-2), 

New  York  and  London. 
MICHELSON,   E.   H.,    1960.      Chemoreception   in   the   snail   Australorbis  glabratus.     Am.    Trap. 

Mcd.  Hyg.,  9:   480-487. 
PAULINI,   E.,   1963.     Field  observations  on  the  upstream  migration  of  Australorbis  glabratus. 

Bull  World  Health  Org.,  29  :  838-841. 
PAWSON,  M.  G.,  1977.     Analysis  of  a  natural  chemical  attractant  for  whiting,  Herlangius  mer- 

langus,  and  the   cod   Gadus   morhua,   using   a   behavioral   bioassay.      Comp.    Biocliein. 

Physiol,,  56A :  129-135. 

PIMENTEL,  D.,  AND  V.   ILDEFONSON,   1957.     Vagility  of  Biomphalaria  glabrata  the  snail  inter- 
mediate host  of  Schistosoma  mansoni  in  Puerto  Rico.     Am.  J .   Trop.  Mcd.  H\g.,  6  : 

576-580. 
PIMENTEL,  D.,   AND  P.   C.   WHITE,   1959.     Biological   environment   and   habits  of  Australorbis 

tjlabratus.      Ecology,  40  :   541-550. 
PIP,  E.,  AND  J.  M.  STEWART,  1976.     The  dynamics  of  two  aquatic  plant-snail  associations.     Can. 

J.  Zoo/.,  54:   1192-1205. 
RADKE,   M.    G.,   ND   L.    S.    RITCHIE,    1961.      Field   observations   on   the    migrations   of   marked 

Australorbis  glabratus  snails.    /.  Parasitol.,  47  :  712. 
RITCHIE,    L.    S.,    1973.      Chemical    control    of    snails.      Pages    458-532    in    N.    Ansari,     Ed., 

Epidemiology  and  control  of  Scliistosoniiasis  (Bilharciasis) .     World  Health  Organisa- 
tion   (ISBN".  3-8055-1340-2),   S.   Karger,   Basel. 

SEPERS,  A.   B.   J.,   1977.     The  utilisation  of  dissolved  organic   compounds   in  aquatic   environ- 
ments.    Hydrobiologia,  52 :  39-54. 
SHIFF,  C.  J.,  AND  B.   GARNETT,   1961.     The  short  term  effects  of  three  molluscicides   on   the 

microflora  and  microfauna  of   small   biologically   stable   ponds   in    Southern   Rhodesia. 

Bull,   World,  Health.   Org..  25:   543-547. 

SODEMAN,  W.  A.,  1973.    The  influence  of  light  on  Biomphalaria  glabrata.    Nautilus,  87  :  103-106. 
SODEMAN,  W.  A.,  AND  M.  C.  DOWDA,  1974.     Behavioural  responses  of  Biomphalaria  glabrata, 

Physiol.  Zoo/.,  47:    198-206. 
SUZUKI,  N.,  AND  D.  TUCKER,  1971.     Amino  acids  as  olfactory  stimuli  in  the  freshwater  catfish, 

Ictalarus  catus  (Linn).     Comp.  Biochm.  Physiol,,  40A  :  399-404. 
THOMAS,  J.  D.,   1973.     Schistosomiasis  and  control  of  molluscan  host  of  human  schistosomes 

with  particular  references  to  self  regulatory  mechanism.     Adv.  Parasitol.,  11  :  307-394. 
TOWNSEND,   C.   R.,   1973a.     The  food-finding  orientation  mechanism  of  Biomphalaria  glabrata 

(Say).  Anim.  Bchav.,  21  :  544-548. 


RHEOTAXIS  IN  filOMPHALAKIA  GLAHRATA  373 

TOWNSEND,  C.  R.,  1973b.     The  role  of  the  osphradium  in  chemoreception  hy  the  snail  Biomp- 

phalaria  ylabrata  (Say).    Aniin.  Bchav.,  21  :  549-556. 
TOWNSEND,  C.  R.,  1974.     The  chemoreceptor  sites  involved  in  food-finding  by  the  freshwater 

snail   Biophalaria   ylabrata    (Say)    with    particular    reference    to   the    function   of   the 

tentacles.    Bchav.  Biol.,  11  :  511-523. 

WATSON,  G.  S.,  1962.     Goodness-of-fit  tests  on  a  circle  II.     Biometrika,  49:   57-63. 
WEBBE,  C.,  AND  P.  JORDAN,   1966.     Recent  advances  in  knowledge  of   Schistosomiasis  in  East 

Africa.     Trans.  R.  Soc.  Trop.  Mcd.  Hyy.,  60  :  279-312. 
WETZEL,  R.  G.,  AND  B.  A.  MANNY,  1972.     Secretion  of  dissolved  organic  carbon  and  nitrogen 

by    aquatic    macrophytes.      Arch.    Hybridal.    Oryan.    Intcrnat.    Vcrcin.    Thcor.    Angc- 

ivandtc.  Limnol..  18:  162-170. 


Reference:  iiwl.  Hull.,  154:  374-382.     (Juiu-.   l'»7S) 


ACID    PHOSPHATASE  DURING  THE  LIFE  CYCLE  OE  THE 
NEMATODE.  PANAGRELLUS  SILUSIAE 

G.  N.  DOERING  AND  E.  E.  PALINCSAR 

Department  of  Hioloi/y,  Loyola   I'tni'crsity  of  Clnctn/o,  t'liicat/o,  Illinois  60626 

Since  the  late  1920's  many  theories  have  been  suggested  as  possible  explanations 
for  aging,  but  little  agreement  seems  to  exist  regarding  its  true  nature. 

Strehler  (1962)  proposed  that  aging  must  be  universal,  occurring  in  all  old 
animals  of  a  species,  and  essentially  absent  in  the  very  young ;  time  dependent, 
progressing  gradually  in  an  individual  and  in  the  population  ;  intrinsic,  due  to  the 
action  of  time  on  the  biological  system,  rather  than  the  result  of  a  pathology  or 
accident ;  and  deleterious,  unfavorably  affecting  the  survival  capacity  of  the  indi- 
vidual organism  in  its  normal  environment. 

Based  on  the  concept  that  aging  is  a  universal  phenomenon  among  metazoans, 
Gershon  (197,0)  considered  the  nematode  to  be  suitable  for  aging  studies,  because: 
first,  it  is  possible  to  obtain  age-synchronized  populations  of  nematodes  and  estab- 
lish survival  curves;  secondly,  the  nematode's  life-span  and  growth  are  not  altered 
by  up  to  9Qr/c  inhibition  of  DNA  synthesis;  and  thirdly,  they  yield  large  popula- 
tions under  easily  controlled  environmental  conditions,  thus  making  them  fit  for 
biochemical  investigations. 

Since  nematodes  are  eutelic  organisms,  cell  division  and  turnover  are  negligible 
and  most  cells  are  already  differentiated  after  hatching  (Hyman,  1951),  making 
any  deteriorative  processes  leading  to  senescence  more  easily  observable.  The 
nematode  chosen  for  this  study  was  the  free-living  form,  Panagrellus  silusiac. 

The  lysosome  has  been  implicated  as  part  of  the  terminal  lytic  aging  process 
(Brock  and  Strehler,  1968;  Heroic!  and  Meadow,  1970;  Hochschild,  1971).  The 
purpose  of  this  study  was  to  investigate  the  lysosomal  enzyme,  acid  phosphatase, 
and  its  isozyme  patterns  in  the  life  of  P.  silusiac.  The  next  step  will  be  to  relate 
these  isozyme  changes  to  a  later  study  centering  on  aging. 

MATERIALS  AND  METHODS 

The  stock  cultures  of  Panagrellus  silusiac  were  maintained  at  23—25°  C.  The 
growth  medium  was  Gerber  Mixed  Cereal,  which  was  mixed  with  distilled  water 
in  a  weight  to  volume  ratio  of  1:5.  Each  culture  was  maintained  for  14  days. 
A  dilute  antibiotic  solution  of  0.6  /xg  penicillin-G  and  10  p.g  streptomycin/ml  was 
added  to  the  nematodes  during  subculturing  to  avoid  contaminating  the  fresh 
cultures. 

Patnu/rcllns  silusiac  is  an  ovoviviparous  animal  with  five  larval  stages.  The 
first  larval  stage  (Li)  is  intrauterine,  but  the  remaining  stages  (  L2,  L3,  L4  and 
adult)  are  free-living.  The  different  stages  were  identified  by  using  the  average 
lengths  of  worms,  based  on  the  method  of  Gysels  and  van  der  Haegen  (1962). 

After  sample  collection,  the  nematodes  were  separated  by  age  using  the  glass 

374 


ACID  PHOSPHATASE  IN  PAX A(^RELLUS  375 

microbead  technique  outlined  by  Sanioiloff  and  Pasternak  (1969).  Since  this 
procedure  only  separated  the  youngest  free-swimming  larval  stage,  it  was  neces- 
sary to  obtain  as  many  L2  worms  as  possible.  Therefore,  seven-day-old  cultures 
were  placed  in  the  dark  for  16  hr  prior  to  sample  collection,  to  induce  the  nenia- 
todes  to  reproduce,  since  P.  silitsisc  tends  to  copulate  more  often  while  in  dark- 
ness, thus  producing  more  L^  larvae. 

When  later  stages  were  studied,  the  L2  larvae  were  allowed  to  molt  at  23- 
25°  C  to  a  more  advanced  stage  of  the  life  cycle.  Following  the  work  of 
Chow  and  Pasternak  (1969),  the  L2  larvae  were  added  to  petri  dishes  containing 
10  ml  of  clear  \%  barley  solution,  so  that  the  ensuing  growth  to  maturation  would 
be  highly  synchronous.  In  the  barley  solution,  L-j  larvae  were  obtained  in  24  hr, 
L4  in  48  hr  and  adults  in  72  hr. 

The  nematodes  were  also  kept  in  the  \%  barley  solution  until  they  were 
10,  15,  20,  and  25  days  old,  in  order  to  study  the  aging  adult  worms.  This 
part  of  the  study  was  conducted  at  5°  C,  which  allowed  the  adults  to  age  at  a 
somewhat  slower  rate  than  normal,  but  not  to  reproduce.  Therefore,  new  L2 
larvae  could  not  be  born  into  the  age-synchronized  cultures.  Every  24  hr,  one  ml 
of  fresh  barley  solution  was  added  to  the  petri  dishes  in  each  experiment,  to  offer 
fresh  nutrients. 

The  nematodes  to  be  studied  were  concentrated  by  centrifugation  and  ground 
with  a  Foredom  tissue  grinder  in  an  ice-cold  container  to  minimize  the  denaturing 
of  the  isozymes.  In  each  experiment,  the  protein  content  of  the  samples  was 
determined  using  the  method  of  Lowry,  Rosebrough,  Farr,  and  Randall  (1951). 
Polyacrylamide  gel  electrophoresis,  based  on  the  methods  of  Davis  (1964)  and 
Ornstein  (1964)  was  used  in  this  study.  The  following  technique  changes  were 
made.  The  bridge  buffer  used  was  an  0.01  M  histidine-NaOH  buffer  of  pH 
7.5,  which  was  suggested  by  Robinson  (1972).  The  gels  were  10%  acrylamide 
and  were  run  at  6°  C  at  4  mAmp/tube.  The  sites  of  acid  phosphatase  on  the 
gels  were  determined  using  the  reaction  method  of  Barka  (1961).  Electro- 
phoretic  mobility  (Ef)  values  were  determined  directly  from  the  gels. 

Densitometric  tracings  of  the  gels  were  made  immediately  after  the  end  of 
incubation,  at  515  nm.  The  relative  activity  of  each  peak  was  calculated  by 
dividing  the  peak  height  by  the  /xg  of  protein  applied  to  the  gel  (Holla,  Weinstein 
and  Lou,  1974). 

Triton  X-100,  a  detergent  which  disrupts  lysosomal  membranes,  was  utilized 
to  determine  the  amount  of  membrane  bound  and  unbound  acid  phosphatase  in 
the  different  nematode  stages.  Modifying  the  procedure  of  Meany,  Gahan  and 
Maggi  (1967),  Triton  X-100  was  added  to  a  mixture  of  the  stages  of  P.  silnsiac 
using  six  different  methods  of  introduction  and  the  following  concentrations : 
0.1 7r,  0.5 f/<,  1.07,  2.5</c,  5.07,  107,  25%,  50';.  and  100^.  The  six  pro- 
cedures for  adding  the  detergent  are  as  follows  :  immediately  before  the  nematodes 
were  ground  for  electrophoresis;  10  min  before  grinding  the  nematodes;  10  min 
at  37°  C  before  grinding  the  tissue  for  electrophoretic  experimentation;  imme- 
diately after  grinding  the  nematodes;  to  pre-ground  tissue  10  min  before  experi- 
mentation; and  to  pre-ground  nematodes  10  min  at  37  C  before  running 
electrophoresis.  Each  concentration  of  Triton  X-100  was  added  to  the  worms 
in  all  six  of  the  procedures.  Each  result  compared  with  the  appropriate  control 


376 


G.  N.  DOERING  AND  E.  E.  PALINCSAK 


of  distilled  water  added  to  the  homogenate,  and  equal  portions  of  nematodes  and 
detergent  were  used  in  each  case. 

To  determine  the  quantity  of  acid  phosphatase  liberated  in  each  trial,  the 
Sigma  total  acid  phosphatase  test  (Sigma  Technical  Bulletin  No.  104,  1963, 
Sigma  Chemical  Company,  St.  Louis,  Missouri)  was  run  on  a  sample  from  each 
test,  at  410  nm  on  a  spectrophotometer.  The  percentage  of  transmittance  was 
converted  to  Sigma  units/ml  of  acid  phosphatase,  using  a  standard  curve  based 
on  para-nitrophenol. 

To  determine  which  structures  within  P.  silusiae  contain  acid  phosphatase,  a 
light  microscopic  study  was  done  on  each  larval  stage.  The  tissue  was  fixed  in 
1  :  10  commercial  formalin  for  1  hr,  and  dehydrated  in  an  ascending  series  of 
ethanols  following  the  procedure  of  Jensen  (1962).  The  worms  were  infiltrated 
with  paraffin,  positioned  in  paraffin  blocks  and  sectioned  at  10  /*.  The  sec- 
tioned tissue  was  affixed  to  glass  slides  and  stained  using  the  acid  phosphatase- 
lead  sulfate  procedure  of  Gomori  (1952),  which  was  modified  by  Jensen  (1956). 
The  control  for  this  study  was  heat-killed  tissue  (i.e.,  boiled  in  distilled  water 
for  5  min)  carried  through  the  entire  staining  process. 

RESULTS 

This  study  showed  that  there  is  a  relationship  between  changes  in  acid 
phosphatase  activity  and  life  cycle  stages  in  Panagrcllns  silusiae.  After  measur- 
ing the  electrophoretic  mobility  (Ef)  values  of  the  stained  bands  on  each  gel,  it  was 
determined  that  ten  separate  and  distinct  isoenzymes  actually  existed.  The  Lo 
and  L3  stages  each  showed  four  isozyme  bands  on  each  gel.  Five  separate  isozymes 
were  present  in  the  L4,  and  six  distinct  bands  were  found  on  the  gels  of  the 
6-day,  10-day,  15-day,  and  20-day-old  nematodes.  The  25-day-old  nematodes 
showed  seven  isozymes  of  acid  phosphatase,  which  was  the  largest  number  present 
in  any  stage  of  the  life  cycle.  Only  two  of  the  ten  isoenzymes  were  present  in 
all  eight  of  the  stages  studied.  The  average  Ef  values  are  listed  in  Table  I. 
Isozyme  10  always  travelled  beyond  the  tracking  dye,  therefore  resulting  in  an 
Ef  value  of  greater  than  1.0. 


Average  electrophoretic  mobility 
that  stage. 


TABLE  I 
values.     A  dashed  line  indicates  an  absence  of  the  isozyme  at 


Band  numbers 

Stage 

1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

L2 

— 

0.1176 

— 

0.2581 

— 

0.3521 

— 

— 

— 

1.1427 

L3 

— 

0.1077 

— 

— 

— 

0.3529 

— 

— 

0.6452 

1.1385 

L4 

— 

0.1176 

— 

— 

— 

0.3516 

0.5156 

— 

0.6464 

1.1406 

6  days 

— 

0.1270 

— 

— 

0.3143 

0.3498 

0.5178 

— 

0.6395 

1.1471 

10  days 

0.0366 

— 

— 

— 

0.3171 

0.3540 

0.5143 

— 

0.6402 

1.1427 

15  days 

0.0342 

— 

0.1370 

— 

0.3093 

0.3501 

— 

—  • 

0.6461 

1.1398 

20  days 

0.0328 

— 

0.1313 

— 

0.3099 

0.3521 

— 

— 

0.6380 

1.1406 

25  days 

0.0351 

— 

0.1405 

— 

0.3170 

0.3513 



0.5956 

0.6453 

1.1385 

ACID  PHOSPHATASE  IN  PANAGRELLUS 


377 


The  isozymes  of  acid  phosphatase  stained  in  one  of  tliree  ways  on  the  gels. 
The  bands  appeared  either  red,  faint  red,  or  yellow  in  color.  By  conducting  a  Sigma 
total  acid  phosphatase  test  on  each  individual  hand,  it  was  determined  that  each 
colored  hand  was  truly  acid  phosphatase.  A  piece  of  blank  gel  was  used  as  a 
standard. 

Densitometric  tracings  were  made  of  the  electrophoretic  gels  from  each  age 
group  (Fig.  1  and  2).  The  relative  activities  of  each  separated  isozymes  were 


L 


TD 


L. 


6  DAYS 


r\ 


YELLOW 


BROMPHENOL  BLUE  (TD) 


FAINT 
RED 

RED 


FIGURE  1.     Densitometric  record  of  acid  phosphatase  in  L2,  L»,  Li,  and  6-day-old  Pana- 
ijrcllus  silusiac. 


378 


G.  N.  DOERING  AND  E.  E.  PALINCSAK 


10   DAYS 


15  DAYS 


20  DAYS 


25  DAYS 


YELLOW 

BROMPHENOL  BLUE 
(TD) 


FAINT 
RED 

RED 


FIGURE  2.  Densitometric  record  of  acid  phosphatase  in  10-day,  15-day,  20-day,  and  25- 
day-old  Panagrellus  silusiac. 

calculated  directly  from  the  tracings.  The  values  obtained  appear  in  Table  IT. 
By  studying  these  values  and  the  densitometric  tracings,  it  can  be  seen  that  the 
individual  isozymes  increase  or  decrease  in  relative  activity  in  correlation  with 
the  life  cycle  stage  of  the  nematode.  However,  as  the  nematode  cycle  progresses, 
there  is  a  general  decrease  in  the  relative  activity  of  the  enzyme  itself,  while 
there  is  a  concurrent  increase  in  the  number  of  isozymes  present. 

The  results  of  the  Triton  X-100  study  are  listed  in  Table  III.  Method  I 
(added  and  ground  immediately)  liberated  no  more  than  2/f  acid  phosphatase 
than  the  distilled  water  control,  while  method  II  (added  and  ground  after  10 
min)  showed  no  increase  in  the  level  of  acid  phosphatase  obtained.  Method  III 
(added  and  ground  after  10  min  at  37°  C)  and  method  IV  (added  immediately 
to  ground  tissue)  yielded  no  more  than  a  3/f  increase  in  acid  phosphatase 
activity.  Method  V  (ground  10  min  after  it  was  added)  showed  no  increase 
in  the  enzyme  level,  and  method  VI  (gound  10  min  after  it  was  added  at  37°  C) 


ACID  PHOSPHATASE  IN  PANAGRELLUS 


379 


TABLE  II 

Relative  activity  of  acid  phosphatase  (peak  height/fig  protein  applied  to  gel), 
an  absence  of  the  isozyme  at  that  stage. 


A  dashed  line  indicates 


Peak  numbers 

1 

2 

3 

4 

5 

6 

7 

8 

9 

10 

L2 

— 

2.50 

— 

2.02 

— 

2.26 

— 

— 

— 

1.07 

L;! 

— 

2.73 

— 

— 

— 

2.32 

— 

— 

0.95 

1.28 

L, 

— 

2.62 

— 

— 

— 

2.62 

1.25 

— 

1.04 

1.18 

6  days 

— 

1.57 

— 

— 

0.67 

0.50 

0.47 

— 

1.11 

0.95 

10  days 

0.22 

— 

— 

— 

0.35 

0.37 

0.33 



1.11 

0.50 

15  days 

0.30 

— 

0.56 

— 

0.35 

0.35 

— 

— 

1.54 

0.37 

20  days 

0.24 

— 

0.38 

— 

0.30 

0.34 

— 

— 

1.04 

0.35 

25  days 

0.22 

— 

0.16 

— 

0.18 

0.30 

— 

0.32 

1.03 

0.26 

liberated  no  more  than  \%  acid  phosphatase  than  the  control.  The  results  of 
these  six  trials  indicate  that  only  a  negligible  amount  of  acid  phosphatase  is 
bound  to  membranes  within  the  cells. 

In  order  to  determine  what  structures  within  Panagrellus  siliisiae  contain 
acid  phosphatase,  the  nematodes  were  specifically  stained  for  the  enzyme,  and 
studied  using  light  microscopy.  The  controls  of  heat-killed  tissue  were  run  for 
each  stage  to  prove  that  any  staining  was  truly  due  to  the  presence  of  acid  phos- 
phatase and  not  to  something  in  the  staining  process  itself.  In  the  experimental 
studies,  any  structure  that  appeared  black  in  color  contained  acid  phosphatase. 

Only  the  digestive  tract  stained  lightly  in  the  L2  stage.  The  Ls  stage  stained 
lightly  throughout  the  length  of  its  body  and  the  faint  line  of  the  intestine 
was  again  visible.  This  indicates  that  these  two  stages  contain  small  amounts 


TABLE  III 

Effect  of  concentration  of  Triton  X-100  on  acid  phosphatase  activities  rising  p-nitrophenol  as  a  sub- 
strate (values  expressed  in  Sigma  units/ml  of  acid  phosphatase). 


Method  used 


Cone,  of 

in 

IV 

V 

VI 

Triton  X-100 

Added  and 
ground 
immediately 

Added  and 
ground 
after  10  min 

Added  and 
ground 
after  10  min 
(37°C) 

Added 
immediately 
to  ground 
tissue 

Ground  1(1 
min  after 
it  was  added 
(37°C) 

Ground  10 
min  after 
it  was  added 
(37°C) 

0.0% 

0.40 

0.34 

0.39 

0.54 

0.45 

0.51 

<>.!% 

0.41 

0.34 

0.34 

0.55 

0.44 

0.45 

0.5% 

0.42 

0.32 

0.38 

0.57 

0.39 

0.44 

1.0% 

0.40 

0.30 

0.39 

0.57 

0.39 

0.50 

2.5% 

0.35 

0.27 

0.41 

0.52 

0.40 

0.50 

5.0% 

0.41 

0.31 

0.42 

0.58 

0.43 

0.52 

10% 

0.35 

0.29 

0.40 

0.49 

0.44 

0.48 

25% 

0.39 

0.27 

0.39 

0.52 

0.44 

0.44 

50% 

0.40 

0.31 

0.34 

0.50 

0.40 

0.47 

100% 

0.35 

0.30 

0.35 

0.55 

0.44 

0.51 

380  G.  N.  DOERING  AND  E.  E.  PALINCSAR 

of  the  enzyme.  In  the  L4  stage,  a  large  amount  of  staining  occurred.  For 
the  first  time,  the  esophagus  and  intestine  stained  darkly,  indicating  that  large 
amounts  of  acid  phosphatase  were  present.  Also,  the  immature  gonads,  which 
begin  to  develop  in  this  stage,  stained  positively  for  the  enzyme.  A  number  of 
structures  stained  in  the  adult  stage,  including  the  entire  gastrointestinal  tract, 
the  fully  developed  reproductive  system,  the  excretory  canals  and  eggs  within  the 
bodies  of  sexually  mature  adult  females.  L!  worms  waiting  to  emerge  from  the 
bodies  of  adult  females  also  stained  lightly,  indicating  the  presence  of  a  small 
amount  of  acid  phosphatase. 

DISCUSSION 

For  the  data  to  be  meaningful,  both  the  unbound  and  membrane-bound  iso- 
zymes  of  acid  phosphatase  have  to  be  considered.  Since  the  unbound  isozymes 
could  be  assayed,  Triton  X-100  was  selected  to  release  those  isozymes  bound  to 
membranes  within  the  cells.  Depending  on  the  method  of  introduction  and  con- 
centration of  the  detergent  used,  no  more  than  39f  of  the  total  acid  phosphatase 
present  in  P.  silusiac  was  found  to  be  bound  by  membranes,  meaning  97(/c  of  the 
enzyme  could  be  assayed  without  the  use  of  Triton  X-100.  Therefore,  the 
amount  of  bound  acid  phosphatase  was  considered  negligible,  and  the  use  of 
Triton  X-100  was  abandoned. 

The  results  of  the  electrophoretic  and  densitometric  studies  indicate  that 
there  is  a  relationship  between  acid  phosphatase  levels  and  specific  life  cycle 
stages  in  Panagrcllus  silusiac.  The  changes  exhibited  by  acid  phosphatase  prob- 
ably result  from  the  involvement  of  several  molecular  subtmits  (isozymes)  in  the 
activity  of  the  enzyme. 

By  studying  Table  II,  it  can  be  seen  that  isozymes  1,  3,  and  5  were  only 
present  in  the  adult  stages,  indicating  they  may  be  connected  with  the  onset  of 
maturity.  Isozyme  2  appeared  only  in  the  larval  stages,  indicating  involvement 
with  the  development  of  the  nematode,  instead  of  the  later  stages.  Because 
isozyme  4  was  only  present  in  the  Lo  stage,  it  seems  to  be  related  to  some  early 
development  in  the  young  larvae.  Isozyme  6  was  present  throughout  the  entire 
life  cycle.  Its  relative  activity  peaked  in  the  L4  stage,  dropped  by  almost  80% 
in  the  molt  to  the  adult  stage,  and  continued  to  drop  during  the  rest  of  the 
life  cycle,  indicating  a  greater  involvement  with  early  development  than  with 
maturation.  Perhaps  it  is  involved  with  the  onset  of  gonadogenesis,  since  its 
activity  peaks  during  the  stage  when  this  process  begins.  The  seventh  isozyme 
was  present  briefly  in  the  middle  of  the  life  cycle,  indicating  that  it  is  involved 
with  the  onset  of  development  of  some  particular  structures  and  disappears 
upon  their  completion.  The  eighth  isozyme  may  be  involved  with  the  final 
aspects  of  aging  since  it  appeared  only  in  the  25-day-old  nematodes.  Isozyme  9 
may  be  connected  with  the  aging  process,  since  its  relative  activity  peaked  in 
the  15-day-old  worms.  The  tenth  isozyme  travelled  beyond  the  tracking  dye  in 
every  stage  of  the  life  cycle.  This  indicates  that  this  isozyme  might  be  of  a  molecular 
weight  less  than  that  of  the  bromphenol  blue  tracking  dye,  and  may  be  involved  in 
both  the  early  and  late  stages,  since  it  is  always  present.  However,  charge  and  other 
factors  involved  in  electrophoresis  might  also  have  caused  such  an  occurrence. 


ACID  PHOSPHATASE  IN  PANAGRELLUS  381 

The  number  of  isozymes  of  acid  phosphatase  increases  during  the  life  span  in 
P.  siliisiac,  but  the  relative  activities  of  the  different  isozymes  peak  at  different 
stages,  while  the  overall  enzyme  activity  decreases  with  maturation.  These  results 
are  consistent  with  the  findings  of  Erlanger  and  Gershon  (1970)  on  T,  accti  and 
Bolla  et  al.  (1974)  on  N.  brasilicnsis,  who  concluded  that  these  biochemical  changes 
correlate  with  the  morphological  and  physical  changes  that  occur  during  the  stages 
of  development  and  aging  throughout  the  nematode's  life  cycle. 

The  isozyme  bands  on  the  electrophoretic  gels  appeared  either  red,  faint  red, 
or  yellow  in  color.  The  difference  between  the  two  shades  of  red  is  explained  by 
the  fact  that  those  isozymes  that  stained  faint  red  were  always  of  a  lesser  activity 
than  those  that  stained  red  on  the  same  gel,  indicating  that  there  was  less  of  each 
faint  staining  isozyme  than  of  those  which  stained  darker.  The  yellow  bands  are 
explained  by  Maclntyre  (1971)  who  studied  the  staining  reactions  used  in  this 
research.  He  found  that  acid  phosphatase  ordinarily  combines  with  two  molecules 
of  fast  garnet,  forming  a  red  dicoupled  colored  complex.  However,  this  usually 
spontaneous  reaction  sometimes  does  not  occur  to  completion,  which  leaves  the  acid 
phosphatase  in  a  yellow,  monocoupled  colored  complex,  with  one  molecule  of  fast 
garnet. 

From  the  light  microscopic  study,  it  can  be  seen  that  the  L2  and  L3  stages  show 
very  little  staining  for  acid  phosphatase,  indicating  a  low  activity  of  the  enzymes. 
However,  the  electrophoretic  and  densitometric  studies  indicated  high  levels  of 
activity  but  few  isozymes  at  these  stages.  Perhaps  some  isozymes  may  not  be 
stained  by  the  histochemical  technique  used.  Due  to  these  somewhat  conflicting 
results,  further  study  of  these  two  larval  stages  is  necessary.  Complete  staining  of 
the  gastrointestinal  tract  and  developing  gonads  was  obvious  in  the  L4  stage.  In 
the  adult  stage,  the  digestive  tract,  excretory  canals,  reproductive  system,  and 
eggs  and  unborn  LI  worms  in  the  bodies  of  mature  females,  all  stained  positively 
for  acid  phosphatase.  This  microscopic  study  therefore  indicates  that  acid  phosphatase 
is  present  in  high  concentrations  in  the  digestive,  excretory,  and  reproductive  sys- 
tems of  Panagrellus  silnsiae,  which  is  consistent  with  previous  findings  in  other 
lower  metazoans  (Cesari,  1974). 

Cristofalo,  Parris  and  Kritchevsky  (1967)  hypothesized  that  with  increased 
age,  acid  phosphatase  activity  gradually  shifts  the  equilibrium  in  the  cell  away  from 
the  synthesis  and  towards  catabolism,  thus  resulting  in  a  general  deterioration  of 
the  cells.  The  data  indicates  that  there  is  a  specific  change  in  the  isozymes  of 
acid  phosphatase  which  corresponds  to  the  stages  in  the  life  cycle.  Acid  phos- 
phatase isozymes  appear  to  vary  with  the  age  of  the  nematodes,  as  discussed  in 
the  Gershon  (1970)  model. 

SUMMARY 

This  study  showed  that  there  is  a  relationship  between  acid  phosphatase  levels 
and  life  cycle  stages  in  the  nematode,  Panagrellus  siliisiac.  Ten  different  isozymes 
of  acid  phosphatase  were  separated  electrophoretically.  Relative  activity  peaked 
at  different  stages  in  the  life  cycle  for  the  different  isozymes.  Later  in  the  life 
cycle,  there  is  a  general  decrease  in  the  relative  activity  of  acid  phosphatase  itself, 
while  there  is  a  concurrent  increase  in  the  number  of  isozymes  present.  At  least 


382  G.  N.  DOERING  AND  E.  E.  PALINCSAR 

97r/c  of  the  acid  phosphatase  in  /'.  sihisiac  is  solubk-  (unbound).  Acid  phosphatase 
appears  to  be  present  in  large  quantities  in  the  entire  gastrointestinal  tract,  the 
excretory  canals,  and  the  reproductive  system  of  mature  Panagrellus  silusiac. 

LITERATURE  CITED 

BAKKA,   T.,   1961.      Studies  of  acid  phosphatase.     I.    Electrophoretic   separation   of   acid   phos- 

phatases  of  rat  liver  on  polyacrylamide  gels.     /.  1  listochcin.  Cytochem.,  9  :  542-547. 
BOLLA,    R.    I.,    P.    P.    WEINSTEIN,   AND   C.    Lou,    1974.      Acid    phosphatase    in   developing   and 

aging  Nippostrongyltts  brasilicnsis.     Comp.  Biochein.  PhysioL,  48B  :   131-145. 
BROCK,  M.  A.,  AND  B.   L.   STREHLKR,   1968.     LUtrastructural  studies  on  the  life  cycle  of  short 

lived  metazoan,  Campanularia  ftexuosa.     J .    I'ltrastntct.  h'cs.,  21  :   281-312. 
CESARI,   I.   M.,    1974.     Schistosoma   nnnisoni:    distribution   and   characteristics   of   alkaline    and 

acid  phosphatase.     Exp.  Parasitol.,  36:  405-414. 

CHOW,   H.    Hu,   AND  J.   PASTERNAK,    1969.      Protein   changes   during   maturation   of   the   free- 
living  nematode,   Panaurellus  silusiae.     J.   Exp.   Zool.,    170 :    77-84. 
CRISTOFALO,  V.   J.,   N.   PARRIS,  AND  D.   KRITCHEVSKY,    1967.      Enzyme  activity   during   growth 

and  aging  of  human  cells  in  vitro.     J .  Cell.  Physiol.,  69  :  263-272. 
DAVIS,    B.    J.,    1964.      Disc   electrophoresis.      II.    Methods    and    applications    to    human    serum 

proteins.    Ann.  N.  Y.  Acad.  Sci.,  121  :  404-427. 
ERLANGER,   M.,  AND  D.   GERSHON,   1970.      Studies  on  aging  in   nematodes.     II.    Studies  of  the 

activities  of  several  enzymes  as  a  function  of  age.     Exp.  Gcrontol..  5  :   13-19. 
GERSHON,  D.,   1970.     Studies  on  aging  in  nematodes.     I.  The  nematode  as  a  model  organism 

for  aging  research.    Exp.  Gcrontol.,  5  :  7-12. 
GOMORI,  G.,  1952.     Microscopic  hist o chemistry,  principles  and  practice.     University  of  Chicago 

Press,  Chicago,  273  pp. 
GYSELS,   H.,   AND   W.   VAN   DER   HAEGEN,    1962.      Post-embryonale   outwikkeling   in   verwilligen 

van    die    vijlevende    nematode    Panagrellus    silusiac     (deMan,     1913),     Goodey,     1945. 

Natuurwet.  Tijdschr.,  44  :  3-20. 
HEROLD,  R.   C.,  AND  N.   D.   MEADOW,   1970.     Age-related  changes  in  infrastructure   and   histo- 

chemistry  of  rotiferan  organs.     /.    Ultnistruct.   Res.,  33  :   203-218. 

HOCHSCHILD,  R.,  1971.     Lysosomes,  membranes  and  aging.     Exp.  Geroutol.,  6:   153-166. 
HYMAN,  L.  H.,  1951.     The  invertebrates:   Vol.  III.     Acanthoccphala,  Aschelininthes  and  E.nto- 

procta.    McGravv  Hill,  New  York,  572  pp. 
JENSEN,   W.   A.,    1956.     The   cytochemical   localization   of   acid    phosphatase    in   root   tip   cells. 

Am.   J.   Botany,  43  :    50-54. 

JENSEN,  W.  A.,   1962.     Botanical  histochemistry.     W.    H.    Freeman  and   Company,    San    Fran- 
cisco, Calif.,  408  pp. 

LOWRY,  O.  H.,  N.  J.  ROSEBROUGH,  A.  L.  FARR,  AND  R.  J.  RANDALL,   1951.     Protein  measure- 
ment with  folin  phenol  reagent.     /.  Biol.  Cliein.,  193  :  265-275. 
MAC!NTYRE,  R.  J.,   1971.     A  method  for  measuring  activities  of  acid   phosphatases   separated 

by  acrylamide  gel  electrophoresis.     Biochein.   Genet.,  5:   45-56. 
MEANY,  A.,  P.  B.  GAHAN,  AND  V.  MAGGI,  1967.     Effects  of  Triton  X-100  on  acid  phosphatases 

with  different   substrate  specificities.     Histochentie,   11  :   280-285. 
ORN STEIN,   L.,    1964.      Disc   electrophoresis.      I.    Background   and   theory.      Ann    N.    Y    Acad 

Sci..   121:  321-349. 
ROBINSON,   H.,   1972.     An  electrophoretic  and  biochemical  analysis  of  acid  phosphatase   in  the 

tail  of  Xcnopiis  laeris  during  development   and  metamorphosis.     /.   Exp.   Zool.     180  : 

127-140. 
SAMOILOFF,  M.  R.,  AND  J.  PASTERNAK,   1969.     Nematode  morphogenesis  :   fine  structure  of  the 

molting  cycles  in  Pamiijrcllits  silusiac    (deMan,    1913),   Goodey,    1945.      Can.   J.   Zool. 

47  :  639-643. 
STREHLER,  B.  L.,  1962.     Time,  cells  and  aging.     Academic  Press,  New  York,  270  pp. 


Reference:  Biol  Bull.,  154:   3X3- 40X.      (June,   l''7X) 


MORPHOLOGY    OF    THE    MOUTHPARTS    OF   LARVAL    LOBSTERS, 

HOMARUS  AM  ERIC  ANUS   (DECAPODA:   NEPHROPIDAE), 

WITH  SPECIAL  EMPHASIS  ON  THEIR  SETAE 

JAN  ROBERT  FACTOR 
Division  of  Biological  Sciences,  Cornell  Unh'crsit\,  Ithacii.  AYii1   York  14853 

The  mouthparts  of  decapod  crustaceans  exhibit  a  rich  diversity  of  form.  Such 
structural  diversity  is  evident  from  the  morphological  descriptions  of  mouthparts 
which  are  generally  included  in  reports  of  research  on  a  variety  of  topics.  Ex- 
amples from  the  Nephropidae  alone  include  morphological  studies  of  adults  (Her- 
rick,  1911),  descriptions  of  larval  development  (Santucci,  1926  and  1927;  Wear, 
1976),  functional  morphology  of  appendages  (Farmer,  1974),  and  descriptions  of 
new  species  (Holthuis,  1974).  Descriptions  of  mouthparts  are  also  featured  in  the 
substantial  literature  on  larval  development  of  non-nephropid  decapods  reared  in 
the  laboratory.  A  small,  but  representative,  sampling  of  such  papers  might 
include  those  of  Conor  and  Conor  (1973a),  who  described  and  illustrated  the 
larval  mouthparts  of  several  porcellanids ;  Bookhout  (1972)  and  Roberts  (1970, 
1973),  who  described  the  mouthparts  of  pagurids ;  and  Costlow  and  Bookhout 
(1959),  Roberts  (1969),  Perkins  (1972).  and  Bookhout  and  Costlow  (1974, 
1977),  who  described  those  of  brachyurans. 

So  far  as  can  be  determined,  no  general  review  of  the  types  of  setae  found  in 
crustaceans  has  been  published ;  in  fact,  few  studies  describing  the  setae  of 
macrurans,  or  of  decapods  generally,  have  been  reported.  Huxley  (1880)  briefly 
mentioned  setae  in  his  classic  work  on  the  crayfish,  Astacus  flnz'iatilis.  More 
recently,  Thomas  described  the  types  and  distribution  of  setae  present  on  the  adult 
(1970)  and  hatchling  stages  (1973)  of  another  British  crayfish,  Austropotamobius 
pallipes,  and  Farmer  (1974)  studied  the  functional  morphology  of  mouthparts  in 
Nephrops  norz'cgicns. 

Surprisingly,  there  are  but  few  such  studies  of  the  American  lobster,  Houiunts 
ainericamts.  While  the  excellent  and  extensive  monographs  of  Herrick  (1896, 
1911)  remain  the  major  works  on  the  morphology  of  this  species,  his  treatment  of 
the  mouthparts  of  larval  lobsters  and  their  setae  is  rather  brief ;  Herrick  left  much 
work  to  be  done  on  these  appendages. 

It  is  the  purpose  of  the  present  work  to  describe  the  types  of  setae  found  on 
the  mouthparts  of  larval  lobsters,  to  devise  a  scheme  for  their  classification,  to 
describe  the  distribution  of  the  various  setal  types,  and  to  present  observations 
of  interesting  or  previously  overlooked  features  of  the  mouthparts  themselves. 
Wherever  possible,  the  structures  described  will  be  considered  in  relation  to  their 
possible  functions.  A  further  purpose  is  to  follow  sequential  changes  in  the 
mouthparts  and  their  setae  which  may  take  place  in  successive  stages.  The  first 
three  stages,  the  larvae,  are  strictly  planktonic,  while  fourth-stage,  postlarval 
lobsters  begin  to  take  up  the  benthic  existence  typical  of  adults.  This  drastic 
change  of  habitat  is  necessarily  accompanied  by  a  change  of  feeding  habits,  and  it 

383 


384  JAN  ROBERT  FACTOR 

seems  inescapable  that  these  changes  should  he  reflected  in  significant  structural 
alterations  in  the  feeding  apparatus.  The  detailed  information  on  the  structure 
of  the  feeding  appendages  presented  in  this  study  may  prove  helpful  in  attempts 
to  devise  a  suitable  artificial  fond  for  the  culture  of  this  commercially  important 
food  species. 

MATERIALS  AND  METHODS 

Collection   of  specimens 

The  lobsters  used  in  this  study  were  collected  at  the  Massachusetts  State  Lob- 
ster Hatchery  and  Research  Station  on  Martha's  Vineyard.  Approximately  forty 
specimens,  hatched  from  several  females,  were  examined. 

Several  external  morphological  features  described  by  Hadley  (1905),  which 
could  easily  be  observed  with  a  dissecting  microscope,  were  used  to  determine  the 
stage  of  each  specimen.  Several  thousand  larvae  are  raised  in  each  rearing  tank 
at  the  Hatchery.  In  the  early  stages  the  intermolt  period  is  relatively  brief,  and 
the  larvae  do  not  molt  synchronously.  It  was  therefore  not  practicable  to  separate 
early  from  late  intermolt  lobsters,  and  no  attempt  was  made  to  identify  stages 
in  the  molt  cycle. 

Procedures  for  light  microscopy 

Specimens  for  light  microscopy  were  fixed  in  seawater  Bouin's  fluid  (Humason, 
1962)  where  they  were  stored  until  needed.  Staining  proved  unnecessary.  Un- 
stained mouthparts  were  mounted  on  microscope  slides  in  an  aqueous  mounting 
medium,  either  Salmon's  polyvinyl  lactophenol  (Type  A. 2)  (Gatenby  and  Beams, 
1950)  or  polyvinyl  pyrrolidone  (Burstone,  1957).  Preparations  were  observed 
with  a  compound  microscope  using  brightfield  and  phase  contrast  illumination. 

Procedures  for  scanning  electron  microscopy 

Specimens  for  scanning  electron  microscopy  were  selected  after  fixation  from 
a  group  of  animals  being  prepared  for  a  future  study  involving  transmission 
electron  microscopy.  The  fixation  procedure  was  thus  more  elaborate  than  that 
customarily  used  for  specimens  to  be  studied  by  scanning  microscopy ;  it  was 
modified  only  slightly  from  the  method  employed  by  Walker  (1976). 

Specimens  were  fixed  in  3%  glutaraldehyde  in  0.2  M  sodium  cacodylate  buffer 
(pH  7.0)  at  6°  C  for  3.5  to  6.5  hours.  The  buffer  contained  a  balanced  salt 
solution  comprising  30  mg/ml  NaCl  and  20  /xg/ml  CaClo  (McDonald,  1972). 
Specimens  were  post-fixed  in  1%  OsO4  in  buffer,  washed  in  solutions  of  buffer 
with  decreasing  concentrations  of  NaCl,  treated  with  2%  uranyl  acetate,  and  de- 
hydrated to  70%:  alcohol  where  they  were  stored.  As  needed,  lobsters  were 
hydrated  in  an  alcohol  series,  dehydrated  in  an  acetone  series,  and  dried  in  a  Sorvall 
critical-point  drier  using  COo  to  replace  acetone.  Mouthparts  were  dissected 
from  whole,  dried  lobsters,  mounted  on  aluminum  stubs  with  Scotch  double-coated 
tape  (No.  666),  and  coated  with  gold/palladium  in  a  Tecknics  Hummer  Coater. 
Preparations  were  observed  with  an  International  Scientific  Instruments  Mini-SEM 
at  an  accelerating  voltage  of  15  kV. 


MOUTHPARTS  OF  LARVAL  LOBSTERS  385 

Preparation  of  line  drawings 

Line  drawings  of  individual  appendages  were  made  by  the  "photo  etching" 
method  described  by  Yob  (1973).  Light  micrographs  of  the  mouthparts  and  setae 
to  be  drawn  were  taken  using  a  compound  microscope  and  were  photographically 
enlarged  to  an  appropriate  scale.  Features  of  the  photographs  to  be  retained  in 
the  drawings  were  outlined  directly  on  the  prints  with  a  fine  Rapidograph  pen. 
When  the  inked  lines  were  thoroughly  dry,  the  background  photographic  images 
were  bleached  with  an  iodine  solution  and  removed  with  fixer,  leaving  only  the 
line  drawings  on  white  photographic  paper. 

RESULTS 

Relative  positions  and  orientation  of  fourth-stage  mouthparts 

The  first  five  pairs  of  appendages  serving  as  mouthparts  (mandibles  through 
second  maxillipeds)  are  flattened  and  make  up  a  series  of  layers  covering  the 
mouth.  The  sixth  pair,  the  third  maxillipeds,  are  not  flattened  but  extend  anteriorly 
to  act  as  grasping  structures.  Since  the  mouthparts  lie  roughly  in  the  frontal 
plane  and  do  not  have  the  same  attitude  as  most  appendages  of  the  lobster,  and 
because  they  are  flattened  and  layered,  the  terms  inner  and  outer  are  most  useful 
for  describing  features  toward  and  away  from  the  mouth,  respectively.  Most 
mouthparts  have  inner  and  outer  surfaces ;  for  example,  the  outer  surface  of  the 
first  maxilla  is  adjacent  to  the  inner  surface  of  the  second  maxilla,  on  the  segment 
just  posterior.  Using  the  terms  inner  and  outer,  as  well  as  medial  (toward  the 
path  of  food)  and  lateral,  and  proximal  and  distal,  it  should  be  possible  to  describe, 
compare,  and  understand  the  positions  of  the  mouthparts  and  the  structures  found 
on  them. 

Types  of  setae 

The  construction  of  categories  of  setae  is  a  useful  aid  in  understanding  the 
variety  of  setal  types  found  on  the  mouthparts  of  larval  lobsters.  Such  a  classi- 
fication is  based  largely  on  the  external  morphology  of  the  setae,  particularly  the 
nature  and  distribution  of  the  setules.  In  the  hope  of  standardizing  terminology, 
Thomas's  (1970)  system  for  naming  the  setae  of  Austropotamobius  pallipes  is 
used  as  the  basis  for  naming  the  groups  of  setae  described  in  the  present  work. 
Although  Thomas's  terms  are  used  when  possible,  changes  and  additions  have 
been  made  whenever  appropriate.  In  their  studies  of  larval  crabs,  Bookhout  and 
Costlow  (1974,  1977)  have  also  followed  Thomas's  terminology  for  naming  setae. 

The  setal  complement  of  the  first  four  stages  of  the  lobster  may  be  arranged 
into  ten  categories.  Most  categories  contain  noticeable  variations  in  the  form  of 
the  setae  assignable  to  them,  emphasizing  the  artificial  nature  of  such  a  classification. 
In  several  categories  the  variation  is  sufficient  to  warrant  subdivision  into  several 
types  (each  designated  by  a  letter  and  a  number)  ;  each  type  may,  however,  be  con- 
sidered a  variation  on  the  general  plan  of  the  category  in  which  it  has  been 
placed.  Within  many  of  the  categories  several  features  of  the  setae  were  incon- 
sistently observable.  Apical  pores,  annulations,  and  bulbous  swellings  of  the 


386 


JAN  ROBERT  FACTOR 


TYPES      OF      SETAE 


A 
PLUMOSE 


Bl  B2 

PAPPOSE 


C2  C3 

PLUMODENTICULATE 


D2 
SERRATE 


Fl  F2 

SERRULATE 


TRISERRATE 


G 
TRISERRULATE 


SIMPLE 


J 
HAMATE 


FIGURE  1.  Diagrammatic  representation  of  the  types  of  setae  found  on  the  mouthparts 
of  larval  lobsters.  (Refer  to  the  text  for  a  detailed  explanation  of  the  setal  types.)  Drawings 
are  not  to  the  same  scale. 

shaft  are  mentioned  only  when  they  are  consistently  or  conspicuously  present.  The 
types  of  setae  encountered  in  the  present  study,  illustrated  in  Figure  1,  will  now 
be  described. 

Plumose,  Type  A.  Plumose  setae  bear  two  distinct  rows  of  long,  fine  setules 
along  most  of  the  length  of  the  shaft.  The  setules  may  be  more  or  less  densely 
arranged,  but  the  rows  are  always  situated  opposite  each  other,  forming  an  angle 
of  180°  (Fig.  19).  These  setae  may  be  segmented  by  constrictions,  or  annulations, 
of  the  outer  surface  of  the  shaft  which  occur  near  the  insertion  on  the  shaft  of  some 
pairs  of  opposed  setules  (Fig.  24). 


MOUTHPARTS  OF  LARVAL  LOBSTERS  387 

Pappose,  Type  Bl.  Typically  pappose  setae  have  long,  fine  sctules  similar  to 
those  of  plumose  setae.  Instead  of  lying  opposite  each  other,  however,  the  setules 
are  loosely  arranged  about  the  shaft  in  a  seemingly  random  manner  (Fig.  15). 

Pappose,  T\pc  HZ.  Densely  pappose  setae  are  similar  to  the  previous  type  but 
with  setules  more  closely  crowded  on  the  shaft.  (These  are  limited  in  distribution 
to  a  single  tuft  on  the  first  maxilla  of  the  fourth  stage.) 

The  setae  included  in  the  plumodenticulate  category,  when  examined  with  the 
light  microscope,  closely  resemble  those  called  plumodenticulate  by  Thomas  (1970), 
but  bear  coarse  or  fine  setules  instead  of  denticulations.  Plumodenticulate  setae 
exihibit  considerable  variation  (Figs.  1,  11). 

Plumodenticulate,  Type  Cl.  The  sparse  setules  of  the  proximal  portion  of  these 
setae,  arranged  in  the  same  manner  as  those  on  pappose  setae,  gradually  give  way  to 
finer  and  more  densely  arranged  distal  setules. 

Plumodenticulate,  Type  C2.  The  sparse  setules  of  the  proximal  half,  arranged 
in  the  same  manner  as  the  setules  on  pappose  setae,  are  sharply  separated  from  the 
finer,  denser  setules  of  the  distal  half.  The  setules  are  similar  to  those  of  type  Cl, 
but  the  transition  from  base  to  tip  is  abrupt.  The  two  regions  may  be  separated  by 
a  bulbous  swelling  of  the  setal  shaft. 

Plumodenticulate,  Type  C3.  Although  there  are  no  setules  on  the  proximal 
half  of  these  setae,  they  have  been  included  in  the  plumodenticulate  category  because 
of  their  resemblance  to  type  C2.  In  fact,  they  would  be  identical  to  type  C2  setae 
if  proximal  setules  were  present.  The  distal  half  of  the  shaft  bears  fine,  densely 
packed  setules.  A  bulbous  swelling  may  lie  midway  along  the  length  of  the  shaft. 

Plumodenticulate,  Type  C4.  This  setal  type  bears  proximal  setules  identical 
to  those  of  type  Cl  and  C2  but  differs  in  having  shorter,  coarser  setules  distally. 
This  type  most  closely  resembles  the  basic  plumodenticulate  setae  as  described  by 
Thomas  (1970). 

All  three  types  of  serrate  setae  are  characterized  by  large,  distinct,  tooth-like 
setules  along  the  distal  half  of  the  shaft.  They  setules  are  clearly  arranged  in  two 
rows  forming  an  angle  of  less  than  180°. 

Sen-ate,  Type  Dl.  These  are  typical  serrate  setae  which  fit  the  general  descrip- 
tion above  and  have  no  additional  setules. 

Serrate,  Type  D2.  In  addition  to  two  rows  of  tooth-like  setules,  this  group 
bears  scale-like  setules  on  the  opposite  side  of  the  shaft  (Figs.  20,  22). 

Serrate,  Type  D3.  A  row  of  shorter,  finer  setules  opposite  the  larger,  tooth-like 
setules  distinguishes  this  group  from  the  other  serrate  setae  (Fig.  18). 

Triscrratc,  Type  E.  Triserrate  setae  bear  three  rows  of  typical  serrate  setules. 
The  setules  of  all  three  rows  are  approximately  equal  in  length  (Fig,  17). 

Serrulate,  Type  Fl.  The  distal  half  of  the  shaft  of  serrulate  setae  appears  to 
bear  denticulations,  or  notches,  when  viewed  with  the  light  microscope.  The 
scanning  electron  microscope  reveals,  however,  that  these  can  be  short,  fine,  peg- 
like  setules,  arranged  in  two  rows  forming  an  angle  of  less  than  180°.  These  setae 


388 


JAN  ROBERT  FACTOR 


are  quite  similar  to  typical  serrate  (Dl)  setae,  but  are  smaller  and  have  shorter, 
finer  setules  (Fig.  16). 

Serrulate,  Type  F2.  Thicker  walls  of  the  shaft  and  a  narrower  lumen  separate 
this  type  of  serrulate  setae  from  type  Fl.  The  subtenninal  pore  is  clearly  visible. 
This  type  appears  to  be  somewhat  similar  to  the  "teazel"  setae  described  by  Thomas 
(1970),  but  the  setules  are  opposite  and  are  not  arranged  in  pappose  fashion.  They 
are  found  only  on  the  first  maxillae  of  the  fourth  stage. 

Triscrrulate,  Type  G.  The  short,  fine,  peg-like  setules  of  the  distal  half  of 
triserrulate  setae  are  arranged  in  three  rows  (Fig.  22).  They  can  be  distinguished 
from  triserrate  (E)  setae  by  their  finer  setules,  and  from  serrulate  (Fl)  setae  by 
the  presence  of  three  rows  of  setules  rather  than  two. 

Cuspidate,  Type  HI.  Cuspidate  setae  are  large,  somewhat  conical,  tooth-like 
setae.  They  are  stout,  with  thick  walls  and  relatively  narrow  lumen,  and  lack 
setules  (Fig.  13). 

Cuspidate,  Type  H2.  These  are  cuspidate  setae,  similar  to  type  HI,  with  sparsely 
arranged,  fine,  short,  almost  needle-like  setules  on  the  shaft  (Fig.  21). 

Simple,  Type  I.  Simple  setae  are  usually  relatively  long  and  thin  and  are 
without  setules  of  any  kind  (Fig.  14).  There  may  be  a  bulb  midway  along  the 
length  of  the  shaft.  Several  much  shorter  simple  setae  have  been  observed  and 
are  included  in  this  category. 

Hamate,  Type  J.  These  are  small,  short  setae  shaped  like  hooks  and  lacking 
setules.  They  were  observed  only  on  the  epipodites  of  the  maxillipeds. 


B 


FIGURE  2.  Mandible  (left)  of  fourth-stage  lobster,  showing  inner  (A)  and  outer  (B) 
surfaces.  Abbreviations  are:  GL,  gnathal  lobe;  MP,  mandibular  palp;  Bl,  pappose  setae; 
Cl  and  C3,  plumodenticulate  setae;  I,  simple  setae;  and  T,  two  major  teeth.  Scale  bar 
represents  0.25  mm.  (Figures  2  and  4-9  are  line  drawings  of  representative  mouthparts  of 
fourth-stage  lobsters  showing  the  types  of  setae  present  and  their  distribution  as  well  as 
the  general  morphology  of  each  mouthpart.  Two  views  of  each  mouthpart  are  illustrated, 
showing  the  inner  and  outer  surfaces;  these  are  not  matched  pairs  of  right  and  left  mouthparts.) 


MOUTHPARTS  OF  LARVAL  LOBSTERS  389 

Detailed  descriptions  of  the  nioiithf>arts 

Because  the  mouthparts  of  the  fourth-stage  lobster  are  more  highly  developed, 
show  more  features  than  corresponding  appendages  of  earlier  stages,  and  presum- 
ably resemble  those  of  the  adult  more  closely,  the  detailed  observations  of  the  fourth 
stage  are  presented  at  the  beginning  of  the  section  on  each  mouthpart.  Observations 
of  the  first  three  stages  then  follow  and  are  compared  to  the  situation  found  in  the 
fourth  stage.  In  this  way  major  developmental  trends  and  interesting  differences 
between  the  earlier  stages  and  the  fourth  stage  can  be  noted.  The  descriptive 
account  is  accompanied  by  a  series  of  line  drawings  of  the  fourth-stage  mouthparts 
(Figs.  2—9)  which  serve  not  only  to  illustrate  the  general  form  of  each  appendage 
but  also  to  map  the  distribution  of  the  various  types  of  setae.  The  drawings  should 
also  be  referred  to  for  details  of  setal  distribution  which  have  not  been  included  in 
the  text. 

Mandibles.  The  most  conspicuous  feature  of  the  mandible  is  a  massive  gnathal 
lobe  at  its  distal  end  (Fig.  2).  The  medial  surface  of  the  gnathal  lobe  is 
responsible  for  the  masticatory  effect  of  the  mandible,  and  it  forms  a  cutting 
and  grasping  edge  consisting  of  a  series  of  teeth.  The  teeth  of  the  fourth-stage 
mandible  are  blunt  and  rounded  and  can  be  seen  with  transmitted  illumination  to  be 
heavily  cuticularized.  Each  tooth  is  solidly  sclerotized  and  lacks  a  lumen.  The 
basal  portion  of  the  cutting  edge  usually  bears  at  least  six  similar  teeth  (Fig.  10). 
Distally,  one  or  two  more  massive  teeth  protrude  past  the  tips  of  the  smaller  teeth. 

The  mandibles  of  fourth-stage  lobsters  are  asymmetrical.  The  single  large 
tooth  of  one  side  fits  between  two  large  teeth  on  the  opposite  mandible  (Fig.  3). 
Distal  to  the  large  teeth  is  a  smaller,  less  heavily  cuticularized  protuberance  which 
is  present  on  both  sides. 

The  mandibular  palp  projects  anteriorly  from  the  base  of  the  gnathal  lobe  on 
the  lateral  surface  of  the  mandible.  The  palp  of  the  fourth-stage  lobster  has  three 
segments ;  the  articulation  between  the  first  and  second  segments,  however,  is 
usually  inconspicuous. 

Almost  all  of  the  setae  present  on  the  mandible  are  found  on  the  two  distal 
segments  of  the  palp.  The  second  segment  bears  several  pappose  (Bl)  and  simple 
(I)  setae,  while  the  terminal  segment  is  heavily  setose,  bearing  type  Cl  and  C3 
plumodenticulate  setae  (Fig.  11).  Several  setae  with  relatively  thick  walls  and  a 
narrow  lumen  were  also  observed  on  some  specimens.  These  are  similar  to  type 
F2  serrulate  setae  but  have  a  terminal,  rather  than  subterminal,  apical  pore.  They 
resemble  even  more  closely  the  setae  Thomas  called  "teazels",  and  they  are  found 
in  the  same  location. 

The  first-stage  mandible  differs  significantly  from  that  of  the  fourth  stage.  The 
teeth  of  the  gnathal  lobe  are  thinner,  sharper,  and  more  delicate,  shaped  like 
slender  cones  (Fig.  12).  The  basal  portion  of  the  cutting  edge  bears  approxi- 
mately ten  to  thirteen  of  these  teeth  with  spaces  between  them.  Each  tooth  has 
a  relatively  narrow,  but  clearly  discernible,  lumen  and  thick,  cuticularized  walls. 
In  this  regard  they  appear  similar  to  some  types  of  setae ;  however,  they  lack  the 
basal  socket  characteristic  of  setae.  At  the  proximal  end  of  the  cutting  edge,  on 
the  inner  surface,  is  a  "pad"  or  dense  field  of  small  setae  which  are  all  directed 
medially,  or  toward  the  mouth  (Fig.  12). 


390 


JAN  ROBERT  FACTOR 


FIGURE  3.  Outer-surface  views  of  representative  matched  pairs  of  right  and  left  mandibles, 
illustrating  the  developmental  changes  occurring  in  first-  through  fourth-stage  lobsters.  All 
drawings  are  to  the  same  scale.  Scale  bar  represents  0.25  mm. 

Unlike  the  chelipeds,  which  are  symmetrical  until  the  fifth  to  eighth  stage 
(Herrick,  1911,  p.  266),  the  mandibles  are  asymmetrical  through  the  fourth 
stage  (Fig.  3).  In  all  specimens  of  the  four  stages  examined  during  this  study, 
the  distal  end  of  the  cutting  edge  on  the  right  mandible  bears  a  single  large  tooth 
and  two  smaller  teeth  associated  with  it  (Fig.  12),  while  the  left  mandible  bears 
two  major  teeth.  It  is  interesting  to  note  that  a  similar  type  of  asymmetry  can 
be  seen  in  the  mandibles  of  adult  lobsters  and  has  been  reported  for  the  adult 
mandibles  of  Ncphrops  nori'Ct/icits  (Farmer,  1974)  and  Austropotamobius  pallipes 
(Thomas,  1970). 

Several  trends  can  be  seen  in  the  mandible  as  lobsters  pass  from  the  first  to 
the  fourth  stage.  First  is  the  tendency  of  the  teeth  on  the  gnathal  lobe  to  become 
more  heavily  cuticularized,  thickening  the  walls  and  decreasing  both  the  size  of 
the  lumen  and  the  space  between  teeth,  until  the  pointed,  comparatively  delicate 
teeth  of  the  first  stage  are  transformed  into  the  massive  molars  of  the  fourth  stage 


MOUTH  1'ARTS  OF  LARVAL  LOBSTERS 


391 


(Fig.  3).  Second-  and  third-stage  mandibles  display  intermediate  conditions, 
although  the  most  drastic  change  occurs  between  the  third  and  fourth  stages. 
The  second  trend  involves  increasing  complexity  of  the  mandibular  palp  (Fig.  3). 
The  first-stage  palp  bears  only  two  setae;  these  are  probably  pappose  (Bl),  but  the 
pattern  of  setules  is  not  obvious.  The  articulations  separating  the  three  segments 
of  the  palp  are  barely  discernible.  In  the  second  stage,  three  setae  are  present, 
clearly  type  C3  plumodenticulate ;  approximately  ten  setae  of  this  type  are  present 
on  the  palp  during  the  third  stage.  By  the  fourth  stage,  the  terminal  segment 
is  covered  by  three  types  of  setae,  as  previously  described.  The  tendency,  then,  is 
an  increase  in  the  number  and  variety  of  setae  present  on  the  mandibular  palp  in 
successive  stages. 

Parac/naths.  Paragnaths  were  noted  in  all  four  larval  ^ages  as  rounded  lobes 
protruding  immediately  posterior  to  the  mouth.  The  surface  facing  the  mouth 
is  covered  by  a  mat  of  extremely  fine,  closely- set,  simple  setae  (Fig.  12).  No 
significant  changes  appear  during  the  larval  stages. 

First  maxillae.  The  first  maxilla  is  composed  of  a  coxal  endite,  a  basal  endite, 
and  an  endopodite  (Fig.  4).  The  medial  edges  of  the  endites  are  the  food-handling 
surfaces,  and  this  is  reflected  in  their  setal  armature.  The  basal  endite  of  the 
fourth  stage  is  provided  with  two  rows  of  stout,  cuspidate  (HI)  setae  along  its 
medial  edge  (Fig.  13),  which  are  obviously  useful  in  manipulating  food.  The  basal 
endite  bears  a  row  of  type  Fl  serrulate  setae  on  its  outer  surface  and  a  row  of  five 
type  F2  serrulate  setae  on  its  inner  surface.  A  clump  of  pappose  (Bl)  setae  lies 
on  the  inner  surface  of  the  basal  endite  near  the  coxal  endite,  and  a  clump  of 
three  setae,  noted  in  several  specimens  to  be  pappose  (Bl),  plumodenticulate  (C4), 
and  serrulate  (Fl),  lies  at  the  anterior  end.  Several  serrulate  (C4)  setae  are 
scattered  on  the  inner  surface  of  this  endite. 

The  medial  edge  of  the  coxal  endite  bears  type  F2  serrulate  setae  distally  and 
typical  pappose  (Bl)  setae  proximally.  Like  the  basal  endite,  the  coxal  endite 


C4 


B 


FIGURE  4.  First  maxilla  (left)  of  fourth-stage  lobster,  showing  inner  (A)  and  outer  (B) 
surfaces.  Abbreviations  are:  BE,  basal  endite;  CE,  coxal  endite;  END,  endopodite;  Bl, 
typical  pappose  setae;  B2,  densely  pappose  setae;  C4,  plumodenticulate  setae;  Fl  and  F2, 
serrulate  setae;  HI,  cuspidate  setae;  and  I,  simple  setae.  Scale  bar  represents  0.25  mm. 


392 


JAN  ROBERT  FACTOR 


A 


FIGURE  5.  Second  maxilla  (left)  of  fourth-stage  lobster,  showing  the  inner  (A)  and 
outer  (B)  surfaces.  (The  number  of  plumose  setae  on  the  scaphognathite  has  been  reduced 
slightly  for  clarity.)  Abbreviations  are:  DBE,  distal  lobe  of  basal  endite;  PBE,  proximal  lobe 
of  basal  endite;  DCE,  distal  lobe  of  coxal  endite;  PCE,  proximal  lobe  of  coxal  endite;  END, 
endopodite;  SCA,  scaphognathite;  A,  plumose  setae;  Bl,  pappose  setae;  C2  and  C4,  plumo- 
denticulate  setae ;  Fl,  serrulate  setae ;  and  I,  simple  setae.  Scale  bar  represents  0.25  mm. 


bears  on  its  outer  surface  a  row  of  type  Fl  serrulate  setae  behind  the  medial  edge. 

The  two-segmented  endopodite  of  the  fourth  stage  bears  typical  pappose  (Bl) 
and  simple  (I)  setae  as  well  as  a  clump  of  densely  pappose  (B2)  setae  near  the 
base  of  the  first  segment.  This  is  the  only  location  where  type  B2  pappose  setae 
were  observed. 

The  first  maxilla  of  the  first  stage  differs  in  several  details.  The  most  interesting 
of  these  involves  the  presence  of  rows  of  small,  fine  setules  on  the  cuspidate  (H2) 
setae  of  the  basal  endite.  These  setules  are  also  present  in  the  second  and  third 
stages.  Here,  as  in  several  other  instances  to  be  noted  below,  the  first,  second, 
and  third  stage  mouthparts  have  cuspidate  setae  with  setules,  while  all  of  the 
cuspidate  setae  observed  on  the  fourth  stage  mouthparts  are  simple. 

The  endopodite  of  this  appendage  in  the  first  three  stages  has  only  a  single 
segment.  The  number  of  setae  at  the  tip  of  the  endopodite  appears  to  undergo  a 
reduction,  as  stages  one,  two,  and  three  have  three  to  five  pappose  (Bl)  and  plumo- 
denticulate  (C2)  setae  in  that  position  instead  of  the  two  found  in  the  fourth  stage. 
The  clump  of  pappose  (B2)  setae,  so  obvious  at  the  base  of  the  fourth-stage  endo- 
podite, is  absent  in  the  first  two  stages  and  represented  by  only  three  setae  in  stage 
three. 

Second  maxillae.    A  bilobed  coxal  endite,  a  bilobed  basal  endite,  an  endopodite, 


MOUTHPARTS  OF  LARVAL  LOBSTERS 


393 


and  the  scaphognathite  constitute  the  second  maxilla  (Fig.  5).  In  the  fourth  stage 
the  medial  edge  of  the  distal  and  proximal  lobes  of  the  basal  endite  bears  mostly 
simple  (I)  and  several  serrulate  (Fl)  setae,  with  one  or  two  pappose  (Bl)  setae 
at  its  distal  end.  The  distal  and  proximal  lobes  of  the  coxal  endite  are  armed  with 
pappose  (Bl),  plumodenticulate  (C4),  serrulate  (Fl),  and  simple  (I)  setae. 

The  one-segmented  endopodite  is  only  sparsely  set  with  setae.  Three  type 
C2  plumodenticulate  setae  are  borne  on  its  tip,  three  pappose  (Bl)  setae  occur 
medially,  and  several  plumose  (A)  setae  are  found  laterally. 

The  scaphognathite  is  a  long,  flat  structure  formed  by  a  fusion  of  the  endo- 
podite (anterior  lobe)  and  epipodite  (posterior  lobe).  It  serves  to  pump  water 
through  the  gill  chamber  and  is  aided  in  this  function  by  a  complete  fringe  of  closely 
set  plumose  (A)  setae. 

Although  the  second  maxilla  follows  the  general  pattern  of  increase  in  size  and 
number  of  setae,  it  appears  to  undergo  less  drastic  changes  than  the  other  mouth- 
parts.  This  may  be  linked  to  the  necessity  for  a  well  developed  scaphognathite 
(for  respiratory  purposes)  from  the  beginning  of  the  lobster's  free-swimming 
existence — the  first  stage.  The  endopodite  of  the  first  stage  bears  pappose  (Bl) 
setae  in  contrast  to  the  types  found  here  in  the  fourth-stage  lobster. 


Bl 


FIGURE  6.  First  maxilliped  (right)  of  fourth-stage  lobster,  showing  the  outer  (A)  and 
inner  (B)  surfaces.  (The  number  of  plumose  setae  on  the  exopodite  has  been  reduced  slightly 
for  clarity.)  Abbreviations  are:  BE,  basal  endite;  CE,  coxal  endite;  END,  endopodite; 
EXO,  exopodite;  EPI,  epipodite;  A,  plumose  setae;  Bl,  pappose  setae;  C2  and  C3,  plumo- 
denticulate setae;  Fl,  serrulate  setae;  and  J,  hamate  setae.  Scale  bar  represents  0.25  mm. 


304 


JAN  ROBERT  FACTOR 


C4 


FIGURE  7.  Second  maxilliped  (right)  of  fourth-stage  lobster,  showing  the  outer  (A) 
and  inner  (B)  surfaces.  Abbreviations  are:  D,  dactyl;  PRO,  propus ;  CAR,  carpus;  MER, 
merus ;  ISC,  ischium;  BAS,  basis;  COX,  coxa;  EXO,  exopodite ;  EPI,  epipodite;  POD,  podo- 
hranch ;  A,  plumose  setae;  Bl,  pappose  setae;  C4,  plumodenticulate  setae;  Dl  and  D2,  serrate 
setae,  I,  simple  setae ;  and  J,  hamate  setae.  Scale  bar  represents  0.25  mm. 


First  maxillipeds.  Basal  and  coxal  endites  are  the  food-handling  structures  of 
the  first  maxillipeds  (Figs.  6,  14).  In  the  fourth  stage  most  of  the  medial  edge  of 
the  hasal  endite  is  occupied  by  type  C2  plumodenticulate  setae  often  with  bulbs 
halfway  along  their  length.  Another  row  of  type  C3  setae  lies  parallel  to  the 
medial  edge  on  the  outer  surface.  A  field  of  serrulate  (Fl)  setae  is  present  on 
the  inner  surfaces  of  the  basal  and  coxal  endites,  and  pappose  (Bl)  setae  fringe 
the  lateral  edge  of  the  basis. 

The  two-segmented  endopodite  bears  a  long  row  of  plumose  (A)  setae  on  its 
inner  surface  in  addition  to  a  row  of  pappose  (Bl)  setae  along  its  medial  and 
lateral  edges. 

The  exopodite  of  the  first  maxilliped  is  divided  into  two  subequal  portions,  a 
basal  segment  and  a  distal  tlagellum  comprising  seven  segments  separated  by  con- 
strictions, or  superficial  folds,  of  the  exoskeleton.  The  lateral  edge  of  the  entire 
exopodite,  the  tip,  and  part  of  the  medial  edge  of  the  flagellum  are  fringed  by 
densely  packed  plumose  (A)  setae  (Fig.  19).  Several  serrulate  (Fl)  setae  are 
found  basally,  near  the  medial  edge. 

The  epipodite  is  a  flattened  structure  which  lies  adjacent  to  the  scaphognathite 
(of  the  second  maxilla)  and  aids  in  the  function  of  moving  water  through  the  gill 


MOUTHPARTS  OF  LARVAL  LOBSTERS  395 

chamber.  Several  serrulate  (Fl)  setae  are  scattered  on  its  outer  surface.  Many 
short,  hooked,  hamate  (J)  setae  are  present  on  most  of  the  surface  of  the  epipodite 
and  form  a  row  along  its  lateral  edge.  Hamate  setae  are  restricted  in  distribution 
to  the  maxillipeds  and,  when  they  are  present,  are  found  only  on  the  epipodites  of 
these  mouthparts. 

The  first  maxilliped  undergoes  relatively  little  change  from  the  first  to  fourth 
stages  besides  an  increase  in  size  and  in  number  of  setae.  The  endopodite  remains 
two-segmented  throughout  the  larval  stages.  In  the  first  stage,  the  exopodite  is 
not  divided  into  two  regions  and  is  fringed  by  plumose  (A)  setae  only  on  its 
lateral  edge;  the  coxa  is  almost  bare  with  only  two  or  three  pappose  (Bl)  setae 
distally.  Hamate  (J)  setae  are  absent  from  the  epipodite  of  the  first  and  second 
stages.  The  third  stage  shows  an  increased  number  of  setae  on  the  coxa,  a 
flagellum  with  three  segments  on  the  exopodite,  and  a  row  of  hamate  setae  on  the 
lateral  edge  of  the  epipodite. 

Second  inaxillipcds.  The  second  maxilliped  is  composed  of  a  protopodite  bear- 
ing an  endopodite,  exopodite,  and  epipodite  (Fig.  7).  The  five-segmented  endo- 
podite, the  food  manipulating  structure,  is  more  elaborate  than  in  any  of  the  pre- 
viously encountered  mouthparts.  The  ischium  and  long  merus  project  anteriorly, 
but  the  carpus,  propus,  and  dactyl  turn  sharply  toward  the  midline  to  give  the 
endopodite  the  shape  of  an  inverted  L. 

The  dactyl  and  propus  are  heavily  armed  with  a  variety  of  setal  types.  The 
tip  of  the  dactyl  bears  two  stout,  cuspidate  (HI)  setae  each  usually  with  a  distinct 
annulation.  Other  setae  present  on  the  dactyl  and  propus  include  serrate  (Dl 
and  D2),  serrulate  (Fl),  and  simple  (I)  types.  Except  for  a  row  of  serrulate 
(Fl)  setae  on  its  inner  surface,  and  a  single  simple  (I)  seta,  the  carpus  is  bare. 

The  merus  is  the  longest  segment  of  the  second  maxilliped,  and  it  bears  almost 
all  of  its  setae  on  the  medial  edge.  Serrulate  (Fl)  and  triserrulate  (G)  setae  are 
present  along  the  entire  medial  edge,  although  the  distal  end  bears  four  or  five 
scaled  serrate  (D2)  setae  (Fig.  20)  and  the  proximal  end  several  plumodenticulate 
(C4)  setae.  Serrulate  (Fl)  and  plumodenticulate  (C2)  setae  are  present  on  the 
small  ischium. 

The  protopodite  is  distinguished  by  the  conspicuous  absence  of  the  coxal  and 
basal  endites  which  are  so  prominent  on  the  maxillae  and  first  maxillipeds.  The 
basis  bears  plumodenticulate  (C4)  and  pappose  (Bl)  setae.  Plumodenticulate 
(C4)  setae  occur  on  the  proximal  portion  of  the  exopodite,  which  extends  from 
the  basis.  The  distal  flagellum  is  distinctly  divided  into  approximately  twelve 
segments  and  bears  segmented  plumose  (A)  setae  near  its  tip. 

Plumodenticulate  (C4)  and  pappose  (Bl)  setae  also  occur  on  the  coxa. 
An  epipodite  extends  from  the  coxa  and  bears  several  scattered  serrulate  (Fl, 
with  distinct  annulations)  and  hamate  (J)  setae.  A  small,  rudimentary  podo- 
branch  protrudes  from  the  base  of  the  epipodite,  near  the  coxa. 

During  the  first  stage,  the  dactyl  of  the  second  maxilliped  bears  at  its  tip  a 
single  cuspidate  (H2)  seta  which  has  needle-like  setules  in  two  rows  along  the 
shaft  (Fig.  21).  The  dactyl  also  bears  two  or  three  serrulate  (Fl)  setae.  Gen- 
erally, fewer  setae  are  present  on  the  second  maxilliped  in  the  first  stage  than  in 
the  fourth.  The  row  of  setae  on  the  inner  surface  of  the  carpus  in  the  fourth  stage 


396 


JAN  ROBERT  FACTOR 


is  represented  in  the  first  stage  by  a  cluster  of  only  three  setae,  and  the  medial 
margin  of  the  merus,  densely  setose  in  the  fourth  stage,  bears  fewer  setae  in  the 
first  stage,  including  serrate  (D2)  setae  with  scales  and  serrulate  (Fl)  and 
triserrulate  (G)  setae  (Fig.  22). 

The  exopodite  is  faintly  divided  into  two  segments  in  the  first  stage,  with  only 
four  or  five  setae  at  its  tip.  Hamate  (J)  and  serrulate  (Fl)  setae  are  not  yet 
present  on  the  epipodite.  Contrary  to  Herrick's  (1896)  observations,  a  rudi- 


D1,D3,G 


D3,G 


C4 


Fl 


FIGURE  8.  Third  maxilliped  (right)  of  fourth-stage  lobster,  showing  the  outer  surface. 
Abbreviations  are:  DAC,  dactyl;  PRO,  propus;  CAR,  carpus;  MER,  merus;  ISC,  ischium; 
BAS,  basis;  COX,  coxa;  EXO,  exopodite;  EPI,  epipodite;  POD,  podobranch;  S,  spines  at 
distal  ends  of  ischium  and  merus;  T,  most  distal  tooth  of  the  row  of  teeth  on  inner  medial 
edge  of  ischium;  A,  plumose  setae;  Bl,  pappose  setae;  C4,  plumodenticulate  setae;  Dl  and  D3, 
serrate  setae;  E,  triserrate  setae;  Fl,  serrulate  setae;  G,  triserrulate  setae;  I,  simple  setae;  and 
J,  hamate  setae.  Scale  bar  represents  0.25  mm. 


MOUTHPARTS  OF  LARVAL  LOBSTERS 

mentary  podobranch  in  the  form  of  a  small  but  distinct  lobe  is  present  near  the 
base  of  the  first-stage  epipodite. 

By  the  third  stage,  the  single  cuspidate  seta  at  the  tip  of  the  dactyl  has  lost  its 
setules.  Also  present  on  the  dactyl  are  several  setae  with  stout  shafts  but  fine 
setules.  It  seems  likely  that  one  of  these  setae  will  lose  its  setules  and  become 
the  second  cuspidate  (HI)  seta,  and  that  others  are  the  forerunners  of  the  serrate 
(Dl)  setae  of  the  fourth  stage.  The  exopodite  has  developed  a  flagellum  of  about 
six  segments  with  plumose  (A)  setae  at  its  tip  and  is  similar  to  the  exopodite  of 
the  fourth  stage. 

The  podobranch  is  somewhat  larger  but  remains  a  simple  lobe  in  the  third 
stage.  Although  the  podobranch  is  still  rudimentary  by  the  fourth  stage,  the  main 
axis  has  acquired  secondary  lobes  typical  of  the  trichobranch  type  of  gill. 

Third  -ina.villipcds.  The  third  maxillipeds,  together  with  the  mandibles,  are  the 
mouthparts  most  responsible  for  the  mastication  of  food.  The  form  of  the  third 
maxillipeds  is  similar  to  that  of  the  second  pair.  A  five-segmented  endopodite, 
flagelliform  exopodite,  and  lamellar  epipodite  extend  from  the  protopodite  (Figs. 
8,  9). 

In  the  fourth  stage,  although  its  lateral  surfaces  are  almost  bare,  the  endopodite 
is  heavily  setose  on  its  medial  surfaces  and  edges,  the  parts  of  the  mouthpart  which 
come  into  contact  with  food  being  passed  to  the  mouth.  The  dactyl  and  propus 
are  flattened  laterally,  so  that  each  has  medial  and  lateral  surfaces  and  inner  and 
outer  edges,  and  bears  serrate  (Dl),  triserrate  (E),  and  simple  (I)  setae. 

The  carpus,  merus,  and  ischium  would  appear  somewhat  triangular  in  cross 
section  and  have  three  edges  which  can  be  considered  inner  medial,  outer  medial, 
and  lateral.  The  edges  define  three  surfaces :  medial,  outer  lateral,  and  inner  lateral 
(Fig.  23).  The  inner  medial  edge  of  the  carpus  is  densely  setose  and  bears 
serrate  (Dl),  triserrate  (E)  (Fig.  17),  and  triserrulate  (G)  setae.  Both  inner 
and  outer  medial  edges  of  the  merus  bear  triserrulate  (G)  and  serrate  (Dl  and 
D3)  setae.  The  distal  end  of  the  lateral  edge  of  the  merus  forms  a  stout  spine 
pointing  distally. 

The  ischium  is  the  longest  segment  of  the  third  maxilliped.  Its  outer  lateral 
surface  bears  short  pappose  (Bl)  setae,  some  of  which  lie  in  a  row  along  the  axis  of 
the  segment.  As  in  the  merus,  the  distal  end  of  the  lateral  edge  of  the  ischium  is 
extended  into  a  substantial  spine  (Fig.  23).  A  row  of  serrate  (D3)  and  triserrulate 
(G)  setae  and  four  or  five  small  spines  pointing  distally  (Fig.  8)  are  present  on 
the  outer  medial  edge.  Of  special  interest  is  the  nature  of  the  inner  medial  edge  of 
the  ischium.  This  edge  consists  of  a  row  of  approximately  fourteen  stout  teeth 
(Fig.  23).  A  row  of  up  to  seven  simple  (I)  setae  is  present  on  the  inner  lateral 
surface,  parallel  to  the  row  of  teeth  (Fig.  23). 

The  medial  edge  of  the  basis  is  an  extension  of  the  outer  medial  edge  of  the 
ischium  and  bears  the  same  kinds  of  spines  and  setae.  The  basis  gives  rise  to  a 
typical  exopodite  divided  into  an  unsegmented  basal  region  and  a  flagellum  of 
approximately  twelve  segments,  most  of  which  bear  a  pair  of  long,  opposite  plumose 
(A)  setae  (Fig.  24).  Plumodenticulate  (C4)  setae  are  scattered  over  much  of 
the  coxa  on  both  its  inner  and  outer  surfaces.  The  epipodite  bears  hamate  (J) 
and  serrulate  (Fl)  setae,  as  well  as  a  row  of  plumodenticulate  (C4)  setae  on  its 


398 


JAN  ROBERT  FACTOR 


FIGURE  9.  Third  maxilliped  (right)  of  fourth-stage  lobster.  View  of  the  inner  surface. 
Segments  and  abbreviations  as  identified  in  Figure  8.  Scale  bar  represents  0.25  mm. 

outer  edge.  Five  rows  of  short  filaments  extend  from  the  central  axis  of  the 
well  developed  podobranch,  which  is  much  further  advanced  than  that  of  the 
second  maxilliped  and  clearly  shows  the  trichobranch  architecture. 

The  structure  of  the  third  maxilliped  in  the  first-stage  larva  is  generally  similar 
to  that  of  the  fourth,  although,  as  in  other  cases,  it  is  less  densely  setose.  Never- 
theless, several  significant  changes  do  occur  between  the  first  and  fourth  stages. 

The  dactyl  bears  only  about  eight  setae  in  the  first  stage,  mostly  of  the  serrulate 
(Fl)  type.  There  are  several  long  setae  at  the  tip  of  the  dactyl,  one  of  which 
is  quite  stout  and  simple  (I)  in  most  specimens.  It  is  reminiscent  of  the  single 
cuspidate  seta  at  the  tip  of  the  first-stage  second  maxilliped  but  is  longer  and 
lacks  thick  walls  and  needle-like  setules.  The  propus  bears  triserrate  (E),  serrulate 


MOUTHPARTS  OF  LARVAL  LOBSTERS  399 

(Fl),  and  serrate  (D2)  setae  with  scales.  On  the  carpus  can  be  found  serrate 
(Dl  and  D2),  serrulate  (Fl)  and  triserrulate  (G)  setae.  The  merus  bears  serrate 
(Dl),  serrulate  (Fl),  and  triserrulate  (G)  setae  and  is  only  indistinctly  separated 
from  the  ischium.  The  setae  on  the  endopodite  are  quite  variable. 

Herrick  (1896,  p.  196)  implies  that  the  prominent  row  of  teeth  on  the  inner 
medial  edge  of  the  ischium  does  not  develop  until  the  fourth  stage.  This  was  found 
to  be  incorrect.  Most  first-stage  specimens  examined  already  have  a  row  of  two 
to  four  rudimentary  teeth.  During  the  second  and  third  stages  approximately  six 
teeth,  rudimentary  but  increasingly  substantial,  are  present.  The  fourth  stage  bears 
roughly  fourteen  formidable  teeth  (Fig.  23). 

Several  serrulate  (Fl)  setae  are  present  on  the  basis  of  the  first  stage.  The 
flagellum  of  the  exopodite  comprises  about  eight  segments,  each  with  a  pair  of 
plumose  (A)  setae.  The  exopodites  on  the  third  maxillipeds  of  the  first  three 
stages  function  as  swimming  organs  along  with  the  exopodites  of  the  five  pereiopods. 
These  are  the  only  instances  in  which  the  exopodites  of  the  mouthparts  aid  in 
swimming. 

Few  serrulate  (Fl)  setae  are  borne  on  the  coxa  in  the  first  stage.  The 
epipodite  may  bear  serrulate  (Fl),  but  no  hamate  (J),  setae  and  has  a  podobranch 
with  three  or  four  rows  of  short  filaments.  It  may  be  noted  that  the  first-stage 
podobranch  of  the  third  maxilliped  is  better  developed  than  the  fourth-stage  podo- 
branch of  the  second  maxilliped.  Several  hamate  (J)  setae  are  present  on  the 
second-stage  epipodite.  The  podobranch  of  the  third  stage  carries  four  rows  of 
filaments  around  the  central  axis;  this  increases  to  five  rows  in  the  fourth  stage. 

DISCUSSION 

Changes  in  the  mouthparts  which  occur  as  lobsters  pass  through  the  larval 
stages  generally  include  increases  in  the  size  of  mouthparts  and  in  the  number  of 
setae  borne  on  them.  Several  other  trends  of  particular  interest,  however,  warrant 
further  discussion. 

It  has  previously  been  mentioned  that  the  mandibles  and  third  maxillipeds  are 
the  most  important  masticatory  appendages.  Observations  of  feeding  behavior  in 
an  adult  lobster  reveal  that  the  teeth  on  the  ischium  of  the  third  maxillipeds  are 
used  to  grasp  one  end  of  a  string  of  food  while  the  other  end  is  held  firmly 
between  the  mandibles.  Downward  movement  of  the  third  maxillipeds  then  causes 
the  food  to  be  stretched  and  torn  as  part  of  the  process  of  mastication  before  it  is 
swallowed. 

The  most  striking  change  in  the  mandible  during  the  larval  stages  is  the 
development  of  the  teeth  of  the  cutting  edge,  in  which  the  relatively  delicate,  seta- 
or  spine-like  teeth  of  the  first  stage  are  progressively  transformed  into  the  molars 
of  the  fourth  stage.  Similarly,  a  prominent  change  in  the  third  maxilliped  is  the 
development  of  the  teeth  on  the  inner  medial  edge  of  the  ischium.  The  few, 
insignificant  teeth  present  in  the  first  stage  are  replaced  by  approximately  fourteen 
substantial  teeth  in  the  fourth  stage. 

Analyses  of  the  stomach  contents  of  hatchery-  and  aquaria-raised  first  through 
fourth  stage  lobsters  by  Williams  (1907)  and  Herrick  (1896)  have  shown  the  diet 
to  comprise  a  variety  of  small  planktonic  organisms,  including  diatoms,  bacteria, 


400 


JAN  ROBERT  FACTOR 


MOUTHPARTS  OF  LARVAL  LOBSTERS  401 

copepods,  filamentous  algae,  and  parts  of  larval  decapods  (including  lobsters). 
Studies  of  stomach  contents  of  adult  lobsters  (Herrick,  1896;  Weiss,  1970;  Ennis, 
1973)  indicate  that  the  diet  includes  crabs,  isopods,  sea  urchins,  sea-stars,  snails, 
clams,  polychaetes,  fish,  eelgrass,  hydroids,  ascidians,  and  ectoprocts.  The  change 
in  diet  which  accompanies  the  lobster's  assumption  of  the  benthic  habitat  requires 
a  corresponding  change  in  mouthparts. 

It  is  evident  that  the  development  of  the  mandibles  and  third  maxillipeds 
enables  lobsters  to  deal  successfully  with  the  more  substantial  food  they  encounter 
in  the  benthic  environment,  which  they  usually  enter  at  the  fourth  or  fifth  stage. 
Furthermore,  the  coordinated  development  of  the  teeth  on  the  mandibles  and  on 
the  ischium  of  the  third  maxilliped  emphasizes  the  coordinated  manner  in  which 
these  appendages  function.  Both  features  appear  necessary  for  the  food-manipulat- 
ing process  typical  of  the  later  stages.  The  development  of  what  appear  to  be 
functional  teeth  on  the  ischium  of  the  third  maxilliped  occurs  at  the  time  when  the 
primary  function  of  these  appendages  changes  from  swimming,  in  the  first  three 
stages,  to  feeding  in  all  subsequent  stages.  The  conclusions  drawn  from  structural 
features  should,  however,  be  augmented  by  observations  of  living,  feeding  larval 
lobsters. 

Larval  lobsters  may  be  usefully  compared  to  the  hatchlings  of  Austropotamobius 
pallipes  described  by  Thomas  (1973).  First-stage  hatchlings  of  this  crayfish  are 
less  well  developed  than  first-stage  lobsters.  They  are  attached  to  the  pleopods 

FIGURE  10.  Scanning  electron  micrograph  of  a  fourth-stage  mandible  (left)  illustrating 
one  of  the  two  major  teeth  and  several  smaller  teeth  of  the  cutting  edge.  Scale  bar  represents 
25  p.  (Figures  10-24  are  scanning  electron  micrographs  illustrating  the  structure  of  the 
mouthparts  and  the  setae  they  bear.) 

FIGURE  11.  Mandibular  palp  of  a  fourth-stage  mandible  (right).  This  view  of  the 
lateral  surface  of  the  terminal  segment  of  the  palp  illustrates  type  Cl  and  C3  plumodenticulate 
setae.  Notice  the  pappose  setules  on  the  proximal  portion  of  the  type  Cl  setae  and  the 
absence  of  setules  on  the  proximal  portion  of  the  C3  setae.  Scale  bar  represents  25  ju. 

FIGURE  12.  Gnathal  lobe  of  a  first-stage  mandible  (right,  outer  surface).  The  distal 
end  of  the  cutting  edge  bears  a  single  major  tooth  and  two  associated  smaller  teeth.  The 
position  of  the  "pad"  of  setae  (P)  can  be  seen  at  the  proximal  end  of  the  cutting  edge, 
although  most  of  the  setae  in  this  field  are  on  the  inner  surface  and  cannot  be  seen  in  this 
view  of  the  outer  surface.  A  paragnath  (PG),  covered  with  very  fine,  simple  setae  can  also 
be  seen.  Scale  bar  represents  50  /t. 

FIGURE  13.  Two  rows  of  cuspidate  (HI)  setae  on  the  medial  edge,  and  a  row  of  ser- 
rulate (Fl)  setae  on  the  outer  surface,  are  visible  on  the  basal  endite  of  a  fourth-stage  first 
maxilla  (right,  outer  surface).  Scale  bar  represents  25  n. 

FIGURE  14.  Distal  (DBE)  and  proximal  (PBE)  lobes  of  the  basal  endite  of  a  first-stage 
second  maxilla  showing  mostly  simple  (I)  setae  on  the  medial  edge.  Scale  bar  represents 
25  M. 

FIGURE  15.  Pappose  (Bl)  setae  from  the  basal  endite  of  a  fourth-stage  first  maxilliped. 
Scale  bar  represents  25  n, 

FIGURE  16.  Serrulate  (Fl)  seta  from  the  propus  and  carpus  of  a  first-stage  second 
maxilliped.  Scale  bar  represents  10  ft. 

FIGURE  17.  A  triserrate  (E)  seta  on  the  carpus  of  a  fourth-stage  third  maxilliped  showing 
three  rows  of  setules  (arrows).  Scale  bar  represents  10  fi. 

FIGURE  18.  A  type  D3  serrate  seta  on  the  outer  medial  edge  of  the  ischium  of  a  fourth- 
stage  third  maxilliped.  The  two  rows  of  tooth-like  setules  can  be  seen  as  well  as  a  third 
row  of  shorter,  finer  setules  (arrow)  which  distinguish  type  D3  from  the  other  serrate  setae. 
Scale  bar  represents  10  M- 


402 


JAN  ROBERT  FACTOR 


MOUTHPARTS  OF  LARVAL  LOBSTERS  403 

of  the  mother  by  a  thread  of  chitin  and  are  not  active  feeders  but  survive  on  the 
remnant  of  the  yolk.  The  mandibles  reflect  this  condition  :  they  have  no  teeth  or 
molar  processes  and  are  lightly  circularized.  First-stage  crayfish  bear  relatively 
few  setae,  and  Thomas  (1970)  suggests  that  those  that  are  present  are  all  asso- 
ciated with  the  creation  of  respiratory  currents,  for  example  the  fringe  of  setae  on 
the  scaphognathite.  In  contrast,  first-stage  lobsters  are  free-swimming,  feeding 
organisms,  with  mouthparts  adequately  developed  to  handle  appropriate  food;  the 
setae  they  bear  are  not  restricted  to  respiratory  current-production  but  presumably 
also  play  an  important  role  in  feeding. 

Conor  and  Conor  (1973),  who  studied  setation  in  procelianid  crabs,  state  that 
analysis  of  the  variations  in  setal  counts  may  be  a  useful  tool  for  distinguishing 
among  different  populations  of  larvae  from  the  same  species.  A  similar  problem 
has  been  approached  by  Rogers,  Cobb  and  Marshall  (1968)  who  used  variations 
in  size  to  distinguish  larvae  of  inshore  populations  from  those  of  offshore  popula- 
tions of  Homants  aincriannts.  The  results  of  Conor  and  Conor  (1973)  suggest 
that  comparisons  of  setation  might  provide  additional  criteria  for  making  this 
distinction. 

While  the  setae  of  decapods  probably  serve  a  variety  of  sensory  functions,  per- 
haps the  most  extensively  studied  is  chemoreception.  Most  research  in  this  area  has 
been  concerned  with  distance  chemoreception  (or  smell )  in  the  antennules,  especially 
the  structure  and  electrophysiology  of  the  aesthetasc  setae  they  bear.  There  has 
also  been  considerable  interest  in  contact  chemoreception  (or  taste)  in  the  setae 
on  the  tips  of  the  pereiopods,  particularly  from  an  electrophysiological  viewpoint. 
Several  studies  of  the  pereiopods  have  also  included  evidence  of  mechanoreception. 
Few  physiological  studies  have  identified  the  setae  acting  as  end  organs  or  have 
even  described  the  types  of  setae  present  in  the  region  under  investigation.  Papers 

FIGURE  19.  Dense  fringe  of  plumose  (A)  setae  on  the  lateral  edge  of  the  exopodite  of  a 
fourth-stage  first  maxilliped.  Scale  bar  represents  25  /j.. 

FIGURE  20.  Second  maxilliped  of  a  fourth-stage  lobster  (right)  showing  four  type  D2 
serrate  setae  at  the  distal  end  of  the  medial  edge  of  the  merus.  Arrows  indicate  the  scale- 
like  setules  typical  of  type  D2  setae.  Scale  bar  represents  25  fj.. 

FIGURE  21.  Tip  of  a  first-stage  second  maxilliped  illustrating  the  single,  terminal  cuspidate 
(H2)  seta  and  the  arrangement  of  serrulate  (Fl)  setae  on  the  dactyl  (DAC)  and  propus 
(PRO).  Arrow  indicates  row  of  setules  on  type  H2  seta.  Scale  bar  represents  50  //.  Inset: 
arrow  indicates  a  single  setule  from  a  type  H2  seta.  Scale  bar  represents  5  /m. 

FIGURE  22.  Several  type  D2  serrate  setae  (from  the  merus  of  a  first-stage  second 
maxilliped)  with  tooth-like  (T)  and  scale-like  (S)  setules.  Also  visible  is  a  triserrulate  (G) 
seta  with  three  rows  of  peg-like  setules  (P).  Scale  bar  represents  10  p.. 

FIGURE  23.  Ischiurn  of  a  fourth-stage  third  maxilliped  (left,  inner  view).  The  row 
of  prominent  teeth  (T)  on  the  inner  medial  edge  of  the  ischium  is  clearly  visible  in  this 
micrograph.  A  row  of  six  simple  (I)  setae  is  present  on  the  inner  lateral  surface  (ILS), 
parallel  to  the  row  of  teeth.  The  distal  end  of  the  ischium  is  extended  into  a  substantial  spine 
(S)  which  overlaps  the  merus  (MER).  The  setae  from  the  outer  medial  edge  are  visible 
behind  the  row  of  teeth.  (Refer  to  Figs.  8  and  9  for  orientation  and  terminology.)  Scale  bar 
represents  100  /JL. 

FIGURE  24.  Plumose  (A)  setae  on  the  tip  of  the  exopodite  of  a  fourth-stage  third 
maxilliped.  Several  annulations  of  the  setal  shaft  (large  arrows)  are  clearly  visible  in  this 
micrograph.  Three  pairs  of  setules  extend  from  each  section  of  the  shaft  between  annulations. 
Small  arrows  indicate  the  points  of  insertion  on  the  shaft  of  intact  setules  or  setules  which 
have  broken.  Scale  bar  represents  10  p. 


404  JAN  ROBERT  FACTOR 

concerned  with  chemoreception  in  Hoinarns  ijannnarns  (  =  H.  vnlyaris}  and 
Hoiiiarus  americanus  include  the  studies  of  Laverack  (1963)  and  Shelton  and 
Laverack  (196cS),  on  the  response  of  chemoreceptors  on  the  dactyl  of  the  walking 
legs  to  various  stimulants,  and  those  of  McLeese  (1970),  Ache  (1972),  Mackie 
(1973),  and  Shepheard  (1974),  who  examined  distance  chemoreception  and  the 
sensitivity  of  chemoreceptors  in  the  antennules. 

Apparently,  however,  very  little  interest  has  been  shown  in  chemosensory  struc- 
tures associated  with  the  mouthparts.  An  exception  is  a  paper  by  Shelton  and 
Laverack  (1970)  who  studied  the  adult  mouthparts  and  pereiopods  of  the  European 
lobster,  Hoinarns  gaiiiinanis.  They  state  that  all  mouthparts  and  pereiopods  bear 
chemoreceptive  endings.  While  these  investigators  made  no  attempt  to  survey 
the  various  types  of  setae  that  might  be  involved  in  chemoreception,  the  type  respon- 
sible for  this  function  appears,  from  their  illustrations,  to  be  identical  with  the 
serrate  setae  described  in  the  present  work. 

It  is  only  to  be  expected  that  many  of  the  setae  on  mouthparts  are  sensory, 
but  the  work  of  Shelton  and  Laverack  (1970)  appears  to  be  the  only  instance  in 
which  this  function  has  been  experimentally  demonstrated.  If  the  serrate  setae 
of  the  larvae  of  Hoinarns  americanus  are  also  chemosensory,  one  would  expect  them 
to  be  located  in  places  where  the  mouthparts  come  into  contact  with  food.  This 
is,  in  fact,  the  case  on  the  second  and  third  maxillipeds  where  the  serrate  setae 
occur  along  the  medial  portions  of  the  appendages.  Although  Shelton  and  Laverack 
(1970)  found  serrate  setae  on  all  six  pairs  of  mouthparts  of  adult  Hoinarns 
yaimnarns,  setae  of  this  type  are  present  only  on  the  second  and  third  maxillipeds 
of  larval  Homarus  americanus.  This  does  not  necessarily  mean  that  chemo- 
reception is  limited  to  the  second  and  third  maxillipeds  among  the  mouthparts ;  it 
is,  more  likely,  an  indication  that  other  types  of  setae  are  also  chemosensory. 

Several  other  functions  of  setae,  in  addition  to  chemoreception  and  mechano- 
reception,  have  been  reported.  Bauer,  studying  the  pandalid  shrimp  Pandalns 
danac  (1975)  and  several  species  of  caridean  shrimps  (1977),  describes  serrate 
setae  on  the  third  maxillipeds  and  pereiopods.  He  assigns  to  these  setae  a  rasp- 
ing function  and  emphasizes  the  role  they  play  in  grooming.  This  position  creates 
a  conflict:  are  the  serrate  setae  rasping  brushes  used  for  grooming  (Bauer) 
or  are  they  chemosensory  end  organs  used  for  tasting  (Shelton  and  Laverack, 
1970)?  Bauer  asserts  (1975,  p.  70)  that  "the  complex  tooth  and  scale  setulation 
are  primarily  adaptations  to  rasping  and  scraping",  although  "all  of  these  serrate 
setae  could  concomitantly  be  chemoreceptive  as  well."  Roberts  (1968)  has  also 
correlated  serrate  setae  with  cleaning  processes  in  his  study  of  hermit  crabs,  and 
Herrick  (1911,  PI.  XXXVI,  Fig.  5)  refers  to  the  medially  directed  groups  of 
setae  on  the  endopodite  of  the  adult  third  maxilliped  as  "cleaning  brushes"  and 
describes  their  use  in  cleaning  the  antennae  of  the  lobster.  It  is  possible  that  the 
serrate  (and  also  serrulate,  triserrate,  and  triserrulate)  setae  on  the  third 
maxillipeds  of  larval  lobsters  aid  in  grooming,  but  observations  of  grooming  behavior 
in  larvae  are  lacking. 

Filter  feeding  has  been  described  in  a  variety  of  decapods.  Conor  and  Conor 
(1973b)  describe  the  changes  which  take  place  in  several  porcellanid  crabs  as 
they  pass  from  the  carnivorous  zoeal  stages  to  the  filter-feeding  megalopa.  Gerlach, 


MOUTHPARTS  OF  LARVAL  LOBSTERS  405 

Ekstrp'm,  and  Eckardt  (1976)  found  that  the  hermit  crab  Pagiirus  bcrnhardus  is 
capable  of  removing  nauplii  of  the  brine  shrimp  (Artcniia)  and  unicellular  algae 
(Dunaliella)  from  suspension.  Budd  and  Lewis  (1977)  report  filter  feeding  in  the 
crayfish  Orconectcs  iininunis  which  appears  to  have  no  elaborate  modifications  of 
the  mouthparts  for  that  purpose.  The  filter  comprises  setae  on  the  second  maxillae 
and  first  maxillipeds.  and  the  exopodites  produce  water  currents  which  carry  par- 
ticles through  the  filter  apparatus.  It  will  be  recalled  that  Herrick  (1896)  and 
Williams  (1907)  found  diatoms  in  the  stomachs  of  larval  lobsters  (Honiarns 
aniericanus)  ;  however,  Williams  (p.  176)  thinks  these  are  not  taken  as  food,  but 
rather  enter  the  stomach  "merely  because  it  is  impossible  to  avoid  these  omni- 
present organisms."  The  possibility  that  diatoms  and  other  very  small  particles 
of  food  are  obtained  by  filtering  the  surrounding  water  deserves  attention. 

Some  setae  present  on  larval  lobsters  are  of  such  form  and  are  present  in  such 
positions  as  to  suggest  that  their  main  function  is  to  aid  in  the  manipulation  of 
food.  Examples  include  the  cuspidate  setae  present  on  the  basal  endite  of  the  first 
maxilla  and  the  cuspidate  and  serrate  setae  on  the  dactyl  of  the  second  maxilliped. 
It  is  easy  to  view  them  as  spikes  which  give  the  mouthparts  purchase  so  that  they 
may  better  move  food  toward  the  mouth,  and  which  at  the  same  time  aid  in 
mastication. 

Still  another  function  of  some  types  of  setae  is  to  extend  the  effective  area  of 
structures  responsible  for  creating  water  currents.  Plumose  setae  seem  best 
adapted  to  this  function.  They  fringe  the  scaphognathite  of  the  second  maxilla, 
which  creates  the  respiratory  current  through  the  gill  chamber,  and  the  exopodites 
of  the  maxillipeds,  which  create  (in  the  adult  lobster,  at  least)  currents  effective 
in  the  removal  of  the  debris  of  feeding.  In  the  first  three  stages  the  exopodites 
of  the  third  maxillipeds,  along  with  those  of  the  pereiopods,  also  serve  the  important 
function  of  swimming.  Additionally,  the  fringe  of  plumose  setae  may  serve  the 
gasket-like  function  of  sealing  the  space  between  the  scaphognathite  and  the  wall 
of  the  gill  chamber,  thereby  preventing  backflow  during  the  pumping  process. 

A  complete  understanding  of  the  functioning  of  the  mouthparts  of  larval  lobsters 
will  require  further  investigation ;  the  first  step  toward  this  end,  however,  is  to 
understand  the  structure  of  the  mouthparts  and  to  appreciate  the  diversity  of  form 
of  the  setae  they  bear. 


I  am  indebted  to  Dr.  John  M.  Anderson  for  directing  my  graduate  education 
and  research,  for  his  extremely  helpful  assistance  in  the  preparation  of  this 
manuscript,  and  for  the  opportunity  to  study  Invertebrate  Zoology  at  Cornell.  I 
am  also  grateful  to  Dr.  M.  V.  Parthasarathy,  for  his  advice  and  careful  training 
in  electronmicroscopical  techniques;  to  Dr.  Lamartine  F.  Hood  (Department  of 
Food  Science,  Cornell),  for  the  use  of  his  scanning  electron  microscope;  to  Dr. 
John  T.  Hughes,  Director  of  the  Massachusetts  State  Lobster  Hatchery  and 
Research  Station,  for  supplying  lobsters  and  allowing  me  to  carry  out  fixations  at 
the  Hatchery ;  and  to  my  good  friend  Dr.  Charles  W.  Walker,  for  his  help,  advice, 
and  encouragement  on  occasions  too  numerous  to  list  here. 

This  paper  represents  part  of  a  thesis  presented  to  the  Graduate  School  of 
Cornell  University  in  partial  fulfillment  of  requirements  for  the  degree  of  Master 


406  JAN  ROBERT  FACTOR 

of  Science.  The  work  was  supported  in  part  by  a  Grant-in-Aid  of  Research  from 
the  Cornell  Chapter  of  Sigma  Xi  and  by  the  Section  of  Botany,  Genetics  and 
Development,  Division  of  Biological  Sciences,  Cornell  University. 

• 
SUMMARY 

1.  This  study  provides  a  detailed  account   of  the  morphology  of   the   mouth- 
parts  of  larval  lobsters  (Homarns  aiucricanus)  and  the  setae  they  bear.     The  results 
describe  the  types  of  setae  found  on  the  motithparts,  present  a  scheme  for  their 
classification,  describe  the  distribution  of  the  various  setal  types,  and  present  obser- 
vations of  interesting  or  previously  overlooked  features  of  the  mouthparts  them- 
selves. 

2.  A  scheme  of  classification   (based  on  the  external  morphology  of  the  setae, 
particularly  the  nature  and  distribution  of  the  setules)  has  been  devised  to  describe 
and  categorize  the  types  of  setae  found  on  the  mouthparts.     The  setal  complement 
may  be  arranged  into  ten  major  categories,  in  some  of  which  the  variation  is  suf- 
ficient to  warrant  subdivision  into  several  types  of  setae. 

3.  Detailed  descriptions  of  the  mouthparts  and  the  distribution  of  their  setae 
are  presented  and  major  developmental  trends  are  noted.     Changes  in  the  mouth- 
parts  which  occur  as  lobsters  pass  through  the  first  four  stages  generally  include 
increases  in  the  size  of  mouthparts  and  in  the  number  of  setae  they  bear.     Of  special 
interest  is  the  transformation  of  the  comparatively  delicate  teeth  of  the  first-stage 
mandible  into  the  massive  molars  of  the  fourth  stage.     The  coordinated  develop- 
ment of  teeth  on  the  gnathal  lobe  of  the  mandible  and  on  the  ischium  of  the  third 
maxilliped  emphasizes  the  coordinated  manner  in   which   these  appendages  func- 
tion.    It  is  evident  that  the  development  of  the  mandibles  and  third  maxillipeds 
enables  lobsters  to  deal  successfully  with  the  more  substantial  food  they  encounter 
in  the  benthic  environment,  which  they  usually  enter  at  the  fourth  or  fifth  stage. 

4.  The  various  functions  and  possible  functions    (for  example,  chemosensory, 
tactile,  and  mechanical )  of  the  setae  borne  on  the  mouthparts  are  discussed  in  light 
of  the  available  functional   and   physiological   evidence.      The   possibility   of   filter 
feeding  in  lobsters,  particularly  in  the  constantly-swimming,  planktonic  larval  forms, 
is    considered. 

LITERATURE  CITED 

ACHE,  B.  W.,  1972.     Ainino  acid  receptors  in  the  antennules  of  /{oiinirus  americanus.     C  <>;;//>. 

Biochcm.  Physiol..  42A  :   807-811. 
BAUER,   R.   T.,   1975.      Grooming   behavior   and   morphology   of  the   caridean   shrimp   Paiulalus 

dunoc  Stimpson   (Decapoda:   Natantia :   Pandalidae).     Zool.  J.  Linn.   Soc.,  56:   45-71. 
BAUER,  R.  T.,   1977.     Antifouling  adaptations  of  marine  shrimp   (Crustacea:   Decapoda:   Cari- 

dea)  :  functional  morphology  and  adaptive   significance  of  antennular  preening  by  the 

third  maxillipeds.    Mar.  Biol,  40  :  261-276. 
BOOKHOUT,   C.   G.,    1972.     Larval   development   of   the   hermit   crab   Pagurus   alatus   Fabricius, 

reared  in  the  laboratory    (Decapoda,   Paguridae).     Crustaceana,  22:   215-238. 
BOOKHOUT,  C.  G.,  AND  J.  D.  COSTLOW,  JR.,  1974.     Larval  development  of  Portitnus  spinicarpus 

reared  in  the  laboratory.    Bull.  Mar.  Sci.,  24  :  20-51. 
BOOKHOUT,  C.   G.,   AND  J.   D.   COSTLOW,  JR.,   1977.     Larval  development   of   Callincctcs  siinilis 

reared  in  the  laboratory.    Bull.  Mar.  Sci.,  27  :  704-728. 
BUDD,  T.  W.,  AND  J.  C.  LEWIS,  1977.     Filter  feeding  in  crayfish.     Atn.  Zonl.,  17  :  921. 


MOUTHPARTS  OK  LARVAL  LOBSTERS  407 

BURSTONE,  M.   S.,   1957.     Polyvinyl  pyrrolidone  as  a  mounting  medium  fur  stains  for  fat  and 

fur  azo-dye  procedures.    Am.  J.  Clin.  Pathol.,  28:  429-430. 
COSTLOW,  J.  D.,  Ju.,  AND  C.  G.  BOOKHOTT,  1959.     The  larval  development  of  Callinectes  sapidus 

Rathbun  reared  in  tlie  laboratory,    Biol.  Hull.,  116  :  373-396. 
Lxxis,   G.   P.,   1973.      Kood,  feeding,  and  condition  of  lobsters,  Homarus  ainericanus,  through- 

out the  seasonal  cycle  in  Bonavista   Bay,   Newfoundland.     /.  Fish.   Res.  Bd.   Can.,  30: 

1905-1909. 
KAKMKU,  A.  S.,  1974.     The  functional  morphology  of  the  mouthparts  and  pereiopods  of  Nephrops 

norrei/iciis  (L.)   (Decapoda:  Nephropidae).    J.Nat.  Hist.  ,8:  121-142. 
GATENBY,  J.   B.,  AND  H.   W.   BEAMS    (Eds.),    1950.      The  Microtomist's   Vade-Mccum    (Bollcs 

Lcc),  llth  Edition.     The  Blakiston  Co.,   Philadelphia,   753  pp. 
GERLACH,   S.   A.,   D.    K.    EKSTR^M,   AND   P.   B.    ECKARDT,    1976.      Filter   feeding   in   the   hermit 

crab,  Pagurus  bernhardus.    Oecologia,  24  :  257-264. 
CONOR,  J.  J.,  AND  S.  L.  CONOR,   1973.     Variations  in  appendage  setal  counts  in  zoea  larvae  of 

four  porcellanid  crabs   (Decapoda  Anomura)   from  Oregon.     Crustaceana.  25:  245-252. 
CONOR,    S.   L.,   AND   J.   J.    CONOR,    1973a.      Descriptions   of   the    larvae   of   four    North    Pacific 

Porcellanidae    (Crustacea:     Anomura).      U.    S.    Fish    M'ildl.    Scrv.    Fish.    Hull.,    71: 

189-223. 
CONOR,   S.   L.,   AND  J.   J.   CONOR,    1973b.      Feeding,   cleaning,   and   swimming  behavior    in   larval 

stages  of   porcellanid   crabs    (Crustacea:     Anomura).      U.   S.   Fish    U'ildl.    Scrv.    Fish. 

null..   71  :  225-234. 
HADLEY,   P.   B.,    1905.      Changes   in   the   form   and   color   in   successive   stages   of  the   American 

lobster    (Homarus   ainericanus).      Preliminary   report.      Rep.   R.    I.    Coiui/i.   Iitl.    Fish., 

35:  44_80. 
HERRICK,   F.   J.,   1896.     The   American  lobster:   a  study  of  its  habits  and   development.     Bull. 

U.  S.  Fish  Cowin.,  15:  1-252. 
HERRICK,   F.  J.,   1911.     Natural  history  of  the  American   lobster.     Bull.    U.  S.   Bitr.   Fish.,   29: 


HoLTiifis,  L.  B.,   1974.     The  lobsters  of  the  Superfamily   Nephropidea  of  the  Atlantic   Ocean 

(Crustacea:  Decapoda).    Bull.  Mar.  Sci..  24:  723-884. 
HTMASON,    G.    L.,    1962.     Animal    tissue    tcchui<iites.      W.    H.    Freeman    and    Co.,    San    Fran- 

cisco, 468   pp. 
HUXLEY,  T.   H.,   1880.      The  crayfish.     An   introduction   to   t/ic  study  of  zoology.     D.  Appleton 

and   Co.,   New   York,   371   pp. 
LAVERACK,   M.   S.,   1963.     Aspects  of  chemoreception  in  Crustacea.     Coin  p.   Biochem.   J'hvsiol. 

8:   141-151. 
MACKIE,  A.  M.,  1973.     The  chemical  basis  of  food  detection  in  the  lobster  Homanis  (jammarus. 

Mar.   Bio!..  21  :    103-108. 
McDoNALD,   K.,   1972.     The  ultrastructure  of  mitosis  in  the  marine   red   alga,  Membranoptera 

platyphylla.    J.  Phycol,  8  :  156-166. 
McLEESE,  D.  W.,  1970.        Detection  of  dissolved  substances  by  the  American  lobster  (Homarus 

americanns]    and   olfactory   attraction   between   lobsters.     /.   Fish.   Res.   Bd.    Can..   27: 

1371-1378. 
PERKINS,  H.  C.,  1972.     The  larval  stages  of  the  deep  sea  red  crab,  Gcryon  ijninqucdens  Smith, 

reared  under  laboratory  conditions    (Decapoda:    Brachyrhyncha).     L'.  S.  Fish   U'ildl. 

Sen:   Fish.   Bull.,   71:   69-82. 
ROBERTS,    M.    H.,    JR.,    1968.      Functional    morphology    of    mouth    parts    of    the    hermit    crabs, 

Paijurus  loni/icarpus  and  Payunis  pollicaris.     Chesapeake  Sci.,  9:  9-20. 
ROBERTS,  M.  H.,  JR.,   1969.     Larval  development  of  Bath\nectes  superha    (Costa)    reared  in  the 

laboratory.     Bi,>l.  Bull.,  137  :  338-351. 
ROBERTS,    M.    H.,   JR.,    1970.      Larval    development   of   Paiiunis   loiii/ienrpus    Say   reared    in   the 

laboratory.      I.   Description  of  larval   instars.     Biol.   Bull..   139:    188-202. 
ROBERTS,   M.   H.,  JR.,   1973.     Larval  development   of  P<i(/i<nts  acatliaiiiis   Benedict,    1901,   reared 

in  the  laboratory   (Decapoda,  Anomura).     Crustaceana,  24:    303-317. 
ROGERS,  B.  A.,  J.  S.  Conn,  AND  N.  MARSHALL,  1968.     Size  comparisons  of  inshore  and  offshore 

larvae  of  the  lobster,  Homarus  americauus,  off  southern   New   England.     Proc.   Natl. 

Shellfish.  Assoc..  58:  78-81. 


40S  JAN  ROBERT  FACTOR 

SANTITCI,    R.,    1926.      Lo    sviluppo   e    1'ecologia    post-embrionali    ddlo    "Scampo"    (Ncphrops 

norvei/icus   (L. ))    nel  Tirreno  e  nei   Mari   Nordici.     Mem.   R.   Com.   Talassogr.  Ital., 

126:   1-36. 
SANTUCCI,  R.,   1927.     Uno  stadio  di  sviluppo  non  ancora  descritto  dello  "Scampo"    (Ncplirops 

norvci/icHs    (L.)).      Mem.   R.    Com.    Talassoijr.    Ital.,    127:    1-7. 
SHELTON,   R.   G.   S.,   AND   M.    S.   LAVERACK,    1968.     Observations   on   a   redescribed   crustacean 

sense  organ.    Comp.  Hiochem.  Pliysiol.,  25  :  1049-1059. 
SHELTON,  R.  G.  S.,  AND  M.   S.  LAVERACK,   1970.     Receptor  bair  structure  and  function  in  tbe 

lobster  Homarus  ;/iimmurus    (L. ).     /.   Exp.   Mar.   Biol.   Ecol.,  4:   201-210. 
SHEPHEARD,   P.,   1974.     Chemoreception  in  the  antennule  of  the  lobster  Homarus  americanus. 

Mar.  Bchav.  Pliysiol..  2  :  261-273. 
THOMAS,  W.  J.,   1970.     Tbe  setae  of  Austropotamobius  pallipes    (Crustacea:   Astacidae).     /. 

Zoo/.  Land.,   160:  91-142. 
THOMAS,  W.  J.,  1973.     The  hatchling  setae  of  Aitstropotumolnus  pullipes   (Lereboullet)    (De- 

capoda,  Astacidae).     Crustaceana,  24:   77-89. 
WALKER,  C.  W.,  1976.     A  comparative  study  of  the  morphology,  histology,  and  ultrastructure  of 

the   reproductive   systems  of  the   sea-stars,   Ctenodiscus   crispatus    (Asteroidea,    Gonio- 

pectinidae)    and    Hippasteria   phryyiana    (Asteroidea,    Goniasteridae) .      Ph.D.    Thesis, 

Cornell   University,   Ithaca,   New   York,   211    pp.      (Diss.   Ahstr..   37B  :    1145-B;    order 

no.  76-21,  136.) 
WEAR,  R.   G.,   1976.     Studies  on  the  larval  development  of  Metanephrops   chaUcngcri   (Balss, 

1914)     (Decapoda,    Nephropidae).      Cnistaccami.   30:    113-122. 
WEISS,  H.   M.,   1970.     Tbe  diet  and  feeding  behavior  of  the  lobster,  Homarus  amerieuuus.  in 

Long  Island  Sound.     Ph.D.   Thesis,  University  of  Connecticut,  Storrs,   104  pp.    (Diss. 

Abstr..  31B:   7245-B ;   order  no.   71-16,057.) 
WILLIAMS,  L.  W.,   1907.     The   stomach  of  the   lobster  and  the  food  of  larval   lobsters.     Rep. 

R.I.   Comm.  In!.  Fish.,  37:    153-180. 
YOB,  P.  C.,  1973.     Photo  etchings.     Petersen's  Photoi/niphic  Magazine,  2  :  32-36. 


Reference:   Biol.  null..  154  :  409-429.     (June,  1978) 


OSMOTIC  AND  IONIC  REGULATION  IN  SEVERAL  WESTERN 

ATLANTIC  CALLIANASSIDAE   (CRUSTACEA, 

DECAPODA,  THALASSINIDEA)1 

DARRYL  L.  FELDER  2 

Department   of  Zoology   and  Physiology,   Louisiana   State    University. 
Baton   Rouge,   Louisiana   70893 

Thalassinid  mud  shrimps  of  the  genera  Callianassa  and  Upogebia  are  frequently 
characterized  as  capable  of  ionic  and  volume  regulation  but  incapable  of  osmotic 
regulation  (Gross,  1957;  Brown  and  Stein,  1960;  Lockwood,  1962;  Kinne,  1963). 
Studies  by  Zenkevich  (1938),  L.  Thompson  and  Pritcharcl  (1969),  and  Hill  (1971), 
however,  document  osmoregulatory  ability  among  upogebids.  The  assumed  ab- 
sence of  this  ability  among  callianassids  is  meanwhile  supported  by  L.  Thompson 
and  Pritchard's  (1969)  studies  of  Callianassa  californiensis  and  C.  filholi.  The 
recent  report  of  strong  ionic  and  osmotic  regulation  in  C.  kraiissi  from  southern 
Africa  (Forbes,  1974)  constitutes  the  first  evidence  of  such  ability  within  the  genus. 
However,  other  Callianassa  species  are  also  in  some  way  adapted  to  low  or  varying 
salinities  (Monod,  1927;  Hedgpeth,  1950;  Wass,  1955;  Phillips,  1971;  Rodrigues, 
1971;  LeLoeuff  and  Intes,  1974).  Generalizations  at  the  generic  level  must, 
therefore,  await  further  physiological  studies  or  perhaps  be  altogether  abandoned 
until  the  systematic  fate  of  the  genus  Callianassa  Leach  has  been  resolved  ;  revisions 
proposed  by  de  Saint  Laurent  (1973),  for  example,  would  partition  Callianassa 
into  six  genera. 

The  present  study  compares  osmotic  adaptations  of  three  species  of  Callianas- 
sidae  from  Louisiana  and  correlates  these  adaptations  to  local  distributions.  Spe- 
cifically, salinity  tolerance,  osmotic  regulation,  and  ionic  regulation  are  reported. 
Despite  the  wide  distribution  of  the  species  concerned,  their  trophic  significance 
(Frankenberg,  Coles,  and  Johannes,  1967),  their  potential  as  bait  fisheries  (Hail- 
stone and  Stephenson,  1961  ;  Bybee,  1969),  and  the  value  of  mud  shrimp  burrows 
in  interpreting  ancient  environments  (Weimer  and  Hoyt,  1964;  Dewindt,  1974), 
basic  understanding  of  their  salinity  tolerances  and  regulatory  capacities  is  lacking. 

Species  concerned  in  the  present  study  are  Callianassa  jainaiccnsc  Schmitt, 
1935,  C.  major  Say,  1818,  and  C.  islagrandc  Schmitt,  1935.  all  of  which  fall  within 
the  subgenus  Callichiriis  Stimpson,  1866.  In  a  study  of  western  African  thalas- 
sinids,  LeLoeuff  and  Intes  (1974)  note  that  Callichinis  is  frequently  euryhaline 
and  typically  restricted  to  littoral  waters  in  tropical  latitudes.  Habitats  of  Callia- 
nassa on  the  Louisiana  coast  are  poorly  documented,  except  in  observations  made 
on  several  coastal  islands  by  Willis  (1942)  ;  he  notes  predominance  of  C.  islagrandc 
on  front  beaches,  interspersion  of  C.  islagrandc  and  C.  major  on  ends  of  islands, 

1  Adapted  from  part   of  a  doctoral  dissertation  submitted   to   the   Department   of  Zoology 
and  Physiology,   Louisiana  State   University,   Baton  Rouge. 

2  Present     address :     Department     of     Biology,  University     of     Southwestern     Louisiana, 
Lafayette,   70504. 

409 


410  DARRYL  L.   FKLDER 

and  predominance  of  ('.  jamaiccnsc  on  back  sides  of  islands  and  in  backbcach 
pools.  Distributions  arc  largely  attributed  to  sediment  characteristics  as  in  a 
later  study  of  C.  islai/ntndc  and  C.  jamaiccnsc  on  the  Mississippi  coast  (Phillips 
1971). 

North  Atlantic  coastal  habitats  of  C.  major  are  described  by  Lunz  (1937),  Pohl 
(1946),  Weimer  and  Hoyt  (1964),  and  Frankenberg  ct  al.  (1967);  limited 
colonization  of  estuary  mouths  is  noted,  and  C.  major  is  usually  reported  from 
higher-salinity  open  beaches.  Rodrigues  (1971)  suggests  some  tolerance  to  varia- 
tions in  salinity  by  C.  major  in  Brazil  but  reports  C.  /amaiccnsc  to  survive  at  the 
mouth  of  the  Rio  Caravelas.  Hedgpeth  (1950)  notes  C.  /amaiccnsc  to  inhabit 
estuarine  mud  Hats  on  the  Texas  coast.  Wass  (1955)  reports  C.  jamaiccnsc 
from  estuaries  in  northwestern  Florida  but  lists  C.  islagrandc  only  from  the  higher- 
salinity  intertidal  zone  of  Gulf  beaches. 

MATERIALS  AND  METHODS 

Studies  were  conducted  from  January,  1972,  to  December,  1974.  Initially, 
distributional  records  were  supplemented  by  collecting  callianassids  from  all  acces- 
sible localities.  Collecting  techniques  included  shoveling  and  sieving,  coring  with 
a  "yabby  pump"  (Hailstone  and  Stephenson,  1961),  and  using  a  portable  water 
jet  to  obtain  specimens,  much  as  described  by  Bybee  (1969).  Except  for  some 
C.  islagrandc  taken  by  shovel  and  sieve,  animals  for  experimental  studies  were 
collected  by  the  water  jet  method  as  it  was  the  most  productive  and  least  injurious 
to  animals. 

All  specimens  of  C.  jamaiccnsc  used  in  experimental  studies  were  collected 
from  the  perimeter  of  a  tidally  influenced  pond  near  the  Louisiana  Wildlife  and 
Fisheries  Commission  Marine  Laboratory  on  Grand  Terre  Island.  To  prevent 
injury  to  animals,  each  was  placed  into  a  perforated,  plastic  vial.  An  insulated 
ice  chest  containing  water  from  the  collecting  site  was  used  to  transport  animals 
to  the  laboratory. 

Animals  were  maintained  unfed  in  individual,  perforated  vials  throughout 
acclimation  periods.  Early  in  the  study,  free-swimming  animals  were  held  in 
sea  water  (SW)  without  isolation,  and  over  90f/f  of  140  C.  jamaiccnsc  perished 
within  two  days  of  collection.  Aggressive  encounters  between  individuals  in  a 
common  container  resulted  in  mutilation  and  consequent  bleeding  which  accounted 
for  high  mortality. 

Within  two  days  of  collection,  after  water  in  the  ice  chest  had  equilibrated  to 
room  temperature  (25  ±  1°  C),  the  animals  and  vials  were  transferred  to  artificial 
SW  equivalent  (±l'/f,  salinity)  to  that  from  the  pond.  Two  to  three  days  were  then 
allowed  for  attrition  of  animals  injured  during  collecting.  Ovigerous  females, 
injured  animals,  and  animals  showing  postmolt  characteristics  detailed  by  L. 
Thompson  and  Pritchard  (1969)  were  not  used  in  experimental  studies. 

Acclimation  solutions  were  prepared  by  dilution  of  artificial  SW  with  deionized 
water.  Salinities  were  approximated  with  a  refractometer.  Animals  were  accli- 
mated stepwise  in  5'/«  increments  or  decrements  per  day  in  the  dark  at  25°  C 
with  continuous  aeration.  Animals  were  maintained  at  the  final  acclimation 
salinity  for  nine  days  before  blood  was  sampled. 


OSMOREGULATION  IN  CALLIANASSIDAE  411 

One  group  of  C.  janiaicense  was  acclimated  to  2()'/((  for  nine  days  after  which 
half  were  weighed  and  transferred  directly  to  3/v,-  ;  the  rest  were  weighed  and 
transferred  to  37/£r.  Animals  were  rinsed  with  deionized  water  and  thoroughly 
blotted  dry  before  being  weighed  to  the  nearest  milligram.  Five  individuals  were 
removed  from  each  salinity  extreme  at  timed  intervals,  rinsed,  blotted  and  re- 
weighed ;  blood  was  then  sampled  and  the  animals  were  lyophilized  to  constant 
weight.  The  same  rinsing  and  blotting  procedures  were  followed  with  all  animals 
from  which  blood  was  sampled. 

Blood  was  obtained  by  puncturing  the  arthrodial  membrane  just  posterior  to 
the  coxa  of  the  fifth  pereiopod  ;  20  /xl  were  drawn  for  determination  of  osmotic 
concentration  and  another  20  /A  were  immediately  diluted  for  ion  analyses; 
squeezing  of  animals  was  avoided.  Osmotic  concentrations  (mOsmol/Kg  H^O) 
of  whole  blood  and  acclimation  media  were  determined  with  a  Hewlett-Packard 
vapor  pressure  osmometer.  An  Aminco  chloride  titrator  was  used  for  chloride 
analyses.  Sodium  was  determined  with  a  Coleman  flame  photometer  and  mag- 
nesium with  a  Perkin-Elmer  atomic  absorption  spectrophotometer. 

Individuals  of  C.  major  used  for  experimental  studies  were  collected  from 
Grand  Isle  and  Grand  Terre  Island  and  those  of  C.  islagrande  were  taken  from 
Isles  Dernieres  and  Chenier  Caminada.  Acclimation  of  these  species  to  salinities 
below  lS(/cc  was  in  some  cases  attempted  in  2.5f/(o  steps.  Collecting,  acclimating, 
blood  sampling,  and  analysis  techniques  were  otherwise  as  described  for  C. 
janiaicense.  Direct  transfers  into  3  and  37 '/((  media  were  not  attempted  with 
C.  major  or  C.  islagrande. 


RESULTS 


Distributions 


As  noted  by  Willis  (1942)  and  Phillips  (1971),  distributions  of  the  species 
studied  are  in  part  determined  by  substrate  characteristics.  Callianassa  janiaicense 
is  found  most  often  in  muddy  substrates  of  back-beach  ponds,  estuarine  flats,  and 
tidal  streams.  Sandier  substrates  of  beaches  facing  the  open  Gulf  are  the  usual 
habitat  of  C.  major  and  C.  islagrandc.  However,  lower  salinities  also  typify  most 
habitats  of  C.  janiaicense;  its  distribution  in  Lousiana  extends  to  well  inside  the 
5'/(,  isohaline  (Chabreck,  1972).  Dense  populations,  with  burrows  exceeding 
200/nr,  are  found  at  2  to  3'/(  salinities  near  Johnson's  Bayou  and  at  5  to  7%c 
in  the  Lafourche  Delta.  On  Grand  Terre  Island  C.  janiaicense  occurs  in  salinities 
which  vary  seasonally  from  6  to  28(/(f,  and  habitat  includes  bayward  margins  of 
Barataria  Pass.  Salinities  at  Barataria  Pass  commonly  change  by  10  to  15'/f 
over  a  period  of  a  few  hours  (Hewatt,  1951). 

By  contrast,  C.  major  and  C.  islagrande  occur  only  in  areas  outside  the  15'^ 
August  isohalines  of  Chabreck  (1972)  and  C.  islagrande  rarely  occurs  inside  the 
20'/f  isohaline.  Callianassa  islgrande  is  the  only  callianassid  found  on  Isles 
Dernieres  front  beaches  which  are  bathed  by  high  salinity  coastal  waters.  Both 
C.  major  and  C.  islagrande  are  found  on  front  beaches  of  Chenier  Caminada  and 
Grand  Isle,  but  C.  major  predominates  on  the  eastern  portion  of  Grand  Isle  where 
salinities  are  less  stable.  On  Grande  Terre  front  beaches,  which  are  inside  the 


412 


I)  \KKY1.   L.   FKI.DKK 


20/v,    isohaline    (Chabreck,    1972),    C.    major   is   abundant    and    C.    islagrande    is 
uncon  11  non. 

During  May,  1975,  a  mixed  population  of  C.  major  and  C.  islagrande  on 
Grand  Isle  was  bathed  by  low  salinity  water  (^7.0/M  for  at  least  four  days. 
Following  the  low  salinities,  numerous  identifiable  decomposing  fragments  of 
C.  islagrande  were  found,  but  only  C.  major  was  found  alive.  Salinity  of  water 
-issuing  from  C.  major  burrows  ranged  from  12  to  \4'/<(.  Of  40  C.  major  collected 
here  half  were  held  at  a  salinity  of  7%n  and  half  were  placed  in  artificial  SW  of  \5%0 
salinity.  Those  held  at  7'/«:  were  dead  within  two  days,  while  most  of  those  in 
\5'/«<  lived  more  than  two  weeks.  Apparently,  substrate  interstitial  water  may 
adequately  buffer  C.  major  from  some  low  overlying  salinities,  while  C.  islagrande 
succumbs  under  the  same  conditions.  Populations  of  C.  islagrande  likely  undergo 
mass  mortalities  where  waters  bathing  beaches  are  subject  to  occasional  extended 
periods  of  low  salinity,  as  may  be  brought  about  by  heavy  rainfall,  high  rates  of 
discharge  from  the  Mississippi  River,  and  the  influence  of  winds  and  tides  on  water 
movement  (Hewatt,  1951).  The  least  vulnerable  populations  of  C.  islagrande 
are  probably  those  on  the  front  beaches  of  Isles  Dernieres,  Timbalier  Island,  and 
the  Chandeleur  Islands  where  salinity  rarely  falls  to  low  levels. 

Mortality,  acclimation,  and  lower  lethal  limits 

All  experiments  were  completed  within  two  to  three  weeks  after  animals  were 
collected.  Of  the  three  species,  C.  jamaicense  proved  most  hardy  in  the  laboratory 
during  and  beyond  this  period,  provided  animals  were  isolated  in  individual  vials. 
Mortality  of  C.  major  and  C.  islagrande  during  the  first  two  to  three  days  after 
collection  ranged  from  7  to  lO'/f,  probably  from  injuries  during  collection;  during 
two  to  three  weeks  thereafter  attrition  ranged  from  2  to  4%  per  week.  Mortality 
of  C.  jamaicense  seldom  exceeded  2%  during  the  first  two  to  three  days  and 

TABLE  I 
Survival  during  attempts  to  acclimate  C.  jamaicense,  C.  major,  and  C.  islagrande  to  low  salinities. 


Species 

Salinity  transfer  (%,) 

Number 
at  Start 

Number  surviving  at  final  salinity 

From 

To 

Step/day 

Day  1 

Day  3 

Day  9 

C.  jamaicense 

2.0 

().() 

2.0 

5 

5 

3 

0 

15.0 

2.0 

5.0  and  3.0 

10 

10 

9 

9 

15.0 

2.5 

5.0  and  2.5 

8 

8 

8 

8 

20.0 

3.0 

17.0 

50 

50 

50 

50 

15.0 

5.0 

5.0 

10 

10 

10 

9 

C.  major 

15.0 

5.0 

5.0 

6 

5 

1 

0 

15.0 

7.5 

5.0  and  2.5 

12 

9 

4 

0 

15.0 

8.0 

5.0  and  2.0 

7 

6 

2 

2 

15.0 

10.0 

2.5 

20 

20 

15 

14 

C.  islagrande 

20.0 

5.0 

5.0 

7 

6 

1* 

1* 

20.0 

10.0 

5.0 

7 

5 

2 

1* 

Juvenile 


OSMOREGULATION  IX  CALLIANASSIDAE 


413 


TAHLK  II 

Changes  in  wet  weight  during  attempts  to  acclimate  C.  major  and  C.  islagrande  to  5  and  JO'^  media. 
(Per  cent  difference  from  original  wet  weight  is  given  as  mean  ±  standard  error  when  more  than  one 
animal  survived;  numbers  in  parentheses  for  mortalities  indicate  numbers  of  animals  surviving  at  the 
end  of  each  day. ) 


Salinitv  transfer 

(  ',,  ) 

Difference  (%)  from  original  wet  weight 

Species 

X  umber  of 

From 

To 

1  day 

2  days 

3  days 

4  days 

9  days 

Mortalities: 

C.  major 

25 

10 

2 

9.  7  ±0.8  (2) 

15.  3  ±4.  6  (2) 

20.9  (1) 

(0) 



10 

5 

5 

15.  5  ±1.3  (5) 

21.  4  ±3.0  (5) 

21.2(1) 

(0) 

— 

C.  islagrande 

15 

10 

4 

14.4±1.4  (4) 

17.7±1.8  (3) 

23.5  ±4.5  (2) 

31.2  (1) 

(0) 

10 

5 

5 

37.6±3.5  (5) 

42.6  (1) 

(0) 

Survivors: 

C.  major 

15 

10 

5 

14.7±1.5 

10.9±1.7 

9.3  ±1.2 

4.8±1.5 

3.  2  ±0.7 

C.  islagrande 

IS 

10 

1* 

13.3 

9.0 

2.7 

0.4 

0.4 

10 

5 

1* 

16.9 

12.3 

11.9 

5.6 

3.2 

*  Juvenile 


averaged  <  \%  per  week  under  stable  conditions  for  up  to  one  month  thereafter. 

Of  ISO  C.  jainaiccnse  specimens  collected  in  March  and  April,  1974,  isolated 
in  vials,  and  held  unfed  in  static  aerated  aquaria  at  25  ±  1°  C,  >  80*%  were  still 
alive  in  November.  Under  similar  conditions,  <  40r/r  of  the  C.  major  and  \7% 
of  the  C.  islagrande  specimens  survived  beyond  two  months. 

Lower  and  upper  lethal  limits  of  salinity  were  apparently  not  reached  when 
C.  jainaiccnse  was  acclimated  to  salinities  from  2  to  45%,  ;  mortalities  were  no 
more  pronounced  at  extremes  of  salinity  than  at  midrange  salinities.  Three  speci- 
mens, transferred  to  deionized  water  after  acclimation  to  2'/«  and  sampling  of  blood, 
survived  in  excess  of  five  days  (Table  I).  No  mortalities  occurred  during  nine  days 
after  direct  transfer  of  50  C.  jainaiccnse  specimens  from  20  to  3'/c,  and  only 
one  animal  died  during  nine  days  after  direct  transfer  of  another  50  from  20  to  37%c. 

The  lower  lethal  salinity  for  C.  major  was  attained  just  below  10%<-  in  several 
acclimation  attempts  (Table  I).  Although  specimens  were  acclimated  to  10%e 
on  several  occasions,  mortalities  during  nine  days  at  the  final  salinity  exceeded 
25%  ;  most  deaths  occurred  during  the  first  three  days  after  the  step  from  12.5  to 
10%c.  Mortalities  for  C.  major  during  acclimation  to  salinities  from  12.5  to  40% r 
did  not  exceed  10%.  In  an  attempt  to  acclimate  five  specimens  to  45%r,  all  animals 
died  between  the  seventh  and  eighth  days  after  transfer  from  40%c. 

Few  specimens  of  C.  islagrande  were  available  for  experimental  studies  and 
tolerance  data  are  preliminary.  Mortality  was  less  than  10%  during  acclimation 
to  salinities  from  20  to  45%<-.  Below  20%f,  acclimation  was  much  less  successful. 
Although  one  C.  islagrande  juvenile  survived  nine  days  at  5%f,  and  another  sur- 
vived nine  days  at  10%f,  all  attempts  to  acclimate  adults  to  salinities  ;  10%r 
resulted  in  100*76  mortalities  within  five  days  (Table  I). 

Weight  changes  were  monitored  during  acclimation  of  some  C.  islayrandc 
and  C.  major  specimens  to  low  salinities.  Mortalities  were  preceded  by  sub- 
stantial increases  in  wet  weights,  which  suggests  inability  to  regulate  volume 
(Table  II).  Moribund  animals  under  low-salinity  stress  had  turgid  abdomens 


414 


DARKY  1.  L.  J-KLDKK 


1200 


Q   1000 

o 

3 

m 

-    300 
en 

UJ 


en     600 
O 


400 


200 


35  % 


200  400  600  800  1000  1200 

MILLIOSMOLES,     MEDIUM 

FIGURE  1.  Blood  osmotic  concentration  in  acclimated  C.  nnijor  (solid  circles)  and 
C.  islayrundc  (open  circles)  as  a  function  of  media  osmotic  concentrations.  Each  large  solid 
circle  is  mean  of  6  to  10  determinations ;  vertical  lines  indicate  range  ;  open  circles  and  small 
solid  circles  are  individual  determinations.  Heavy  line  is  fitted  to  points  for  C.  innjur. 
Asterisk  denotes  juveniles. 

and  branchiostegites  which,  by  restricting  movement  and  ventilation,  probably 
caused  respiratory  stress.  Those  which  survived  low-salinity  acclimation  increased 
in  weight  initially,  but  began  to  regulate  volume  within  three  days;  by  the  fourth 
day  to  ninth  day  of  acclimation,  wet  weights  returned  to  near  original  values. 


Osmotic  and   ionic   regulation 

Blood  of  C.  major  and  C.  islagrandc  is  nearly  isosmotic  to  media  over  the 
entire  salinity  range  in  which  these  animals  survive  (Fig.  1).  Slightly  hyper- 
osmotic  values  obtained  for  C.  major  at  S'/,<  salinity  represent  a  low  percentage 
of  animals  which  survived  at  that  extreme.  Likewise,  hyperosmotic  values  for 
C.  islagrandc  at  low  salinities  are  from  two  juveniles  which  survived  while  nine 
adults  died  at  these  salinities. 

Blood  of  C.  jainaiccnsc  is  hyperosmotically  regulated  at  salinities  <  20'/,f  and 
shows  little  depression  of  osmolality  at  the  lowest  salinity  extreme  of  2'/< ,  (Fig.  2). 
Slightly  higher  levels  of  hvperosmicity  are  maintained  by  January  animals  col- 


lected from  a  field  salinity  of  \\'/«  and  temperature  of  8°  C  than  by  August  animals 


collected   from   23'/<, 


and    33°    C.      Blood    is    isosmotic    to    most    media    salinities 


>  25'/f,f>  and  very  slightly  hyposmotic  at  the  upper-salinity  extreme  of  45'/< 


OSMOREGULATION  IN  CALLIANASSIDAE 


415 


Blood  chloride  is  hypoionic  in  C.  major  and  adult  C.  islagrandc  over  the  entire 
range  of  acclimation  salinities,  though  less  so  at  lower  salinities  (Fig.  3). 
Blood  chloride  was  hyperionic  in  the  two  juveniles  of  C.  islagrandc  which  sur- 
vived acclimation  to  5  and  \0'/(,  salinities.  In  C.  jamaiccnsc  hlood  chloride  is 
hypoionically  regulated  at  salinities  >20'/(t,  isoionic  at  15'/,,  and  hyperionic  at 
salinities  <  l5'/tf  (Fig.  3). 

At  salinities  <  20'/(  hlood  sodium  in  C.  major  is  isoionic  to  acclimation  media 
(Fig.  4).  Hypoionic  regulation  of  sodium  is  exhibited  at  higher  salinities,  hut 
not  to  levels  as  low  as  chloride.  As  with  osmolality  and  chloride,  sodium  is  hyper- 
regulated  in  the  C.  islagrandc  juveniles  surviving  low-salinity  acclimation  (Fig. 
4).  In  C.  islagrandc  adults  acclimated  to  salinities  from  25  to  45'/ff  blood  sodium 
is  hypoionically  regulated  to  levels  approximating  those  for  C.  major.  Sodium 
levels  in  acclimated  C.  jamaiccnsc  are  slightly  hypoionic  to  media  at  salinities 
>  25'/,(,  near  isoionic  at  20/£f,  and  markedly  hyperionic  at  lower  salinities  from 
15  to  2'/tt  (Fig.  5). 

Blood  magnesium  in  acclimated  C.  major  is  to  some  degree  hyper-regulated  at 
salinities  <  30'/r  (Fig.  6)  ;  hlood  concentrations  are  maintained  at  about  6.0  mM/ 
liter  higher  than  media  concentrations  in  salinities  from  X  to  2Q(/(f.  Hyper-regula- 
tion of  magnesium  is  diminished  at  W/,  and  concentrations  fall  to  slightly  hypoionic 
levels  at  37(/«.  In  acclimated  C.  jamaiccnsc,  magnesium  is  slightly  hyper-regulated 


o 
o 


1400  h 


1200 


1000 


m 


-     800 


<r>     600 
O 


400 


200 


35  %, 


200  400  600  800  1000 

MILLIOSMOLES,     MEDIUM 


1200 


1400 


FIGURE  2.  Blood  osmotic  concentration  in  C.  jamaicense  as  a  function  of  media  osmotic 
concentrations  after  acclimation  of  summer  (open  circles)  and  winter  (solid  circles)  animals 
at  a  media  temperature  of  25°  C.  Each  open  circle  is  mean  of  5  to  6  determinations,  and 
each  solid  circle  is  mean  of  8  to  10  determinations ;  vertical  lines  indicate  range. 


416 


DARRYL  L.  FELDER 


600 


§   500 


DO 
a: 

LJ 


400 


E   300 


LJ 
Q 

or 
o 


o 


200 


100 


100  200  300  400  500 

CHLORIDE     mM/LITER     MEDIUM 


600 


FIGURE  3.  Blood  chloride  concentration  in  acclimated  C.  jainaiccnsc  (solid  squares), 
C.  uiajor  (solid  circles)  and  C.  islagrande  (open  circles)  as  a  function  of  media  chloride 
concentrations.  Each  solid  square  or  large  solid  circle  is  mean  of  6  to  10  determinations  ; 
each  small  solid  or  open  circle  is  value  for  individual  animal  ;  vertical  lines  indicate  range. 
Asterisk  denotes  juveniles. 


at   salinities  from    10   to   25'/c    and   more   markedly   hyper-regulated   below 
(Fig.  6).     Slight  hypoionic  regulation  is  exhibited  by  C.  jainaiccnsc  at  salinities 
Blood  magnesium  for  C.  islagrande  was  not  determined. 


Osniorcgiilatory  response  of  C.  jamaicense  to  dramatic  salinity  changes 

Temporal  changes  in  body  water,  blood  osmolality,  and  blood  ion  concentration 
were  monitored  following  direct  transfer  of  20^f-acclimated  animals  to  salinities  of 
either  37(/((>  or  Z'/cc,.  Body  water  increases  slightly  (  —  I'/f  )  during  the  first  12 
hours  after  direct  transfer  to  3(/fc  medium  but  is  maintained  at  levels  equal  to  or 
slightly  less  than  original  values  after  1  day  (Fig.  7).  Osmotic,  chloride,  and 
sodium  concentrations  of  blood  fall  to  near  or  just  below  stable  concentrations 
during  the  first  12  hours  in  3'/ti  medium  (Figs.  8  and  9).  Osmotic  and  sodium 
concentrations  of  blood  show  a  slight  undershoot  after  12-24  hours,  but  at  no 
time  fall  to  the  levels  of  the  medium.  Blood  magnesium  levels  drop  little  over  the 
first  6  to  12  hours,  briefly  recover,  and  then  continue  to  drop  at  a  decreasing  rate 
over  the  entire  nine  day  period  (Fig.  10).  Near  stable  levels  of  blood  magnesium 
are  achieved  after  four  days,  and  concentrations  are  maintained  above  that  of 
the  3'/«  medium. 

When  C.  jamaicense  is  directly  transferred  to  37(/K   medium,  body  water  de- 


OSMOREGULATION  IN  CALLIANASSIDAE 


417 


creases  by  — 3%  over  the  first  six  hours  and  remains  below  original  levels  until 
the  second  day  (Fig.  7).  Osmotic,  chloride,  and  sodium  concentrations  of  blood 
increase  markedly  during  the  first  day  and  continue  to  increase,  at  a  decreasing 
rate,  through  day  9  (Figs.  8  and  9)  ;  near  stable  levels  are  attained  by  day  4.  The 
levels  of  blood  osmolality  and  sodium  on  clay  9  approximate  those  of  the  37/< 
medium.  Blood  chloride  remains  slightly  hypoionic  to  chloride  concentrations  of 
the  medium  through  day  9.  Blood  magnesium  increases  slowly  until  the  fourth  day 
when  it  stabilizes  at  a  level  just  below  that  of  the  medium  (Fig.  10). 

DISCUSSION 

Investigations  by  Teal  (1958),  Snelling  (1959),  Kinne  (1963),  and  Barnes 
(1967)  are  among  those  which  correlate  osmoregulatory  capacities  of  decapod 
crustaceans  to  their  differential  penetration  of  estuaries.  Distributions  of  callianas- 
sids  on  the  Louisiana  coast  also  correlate  with  their  osmoregulatory  capacities  and 
tolerance  of  dilute  media.  This  is  not  to  say  that  habitat  preference  is  solely 
or  even  primarily  determined  by  salinities.  For  example,  despite  its  survival 
in  varying  salinities,  Einerita  talpoida  is  localized  on  wave-washed  beaches  by  its 
feeding  specialization  (Bursey  and  Bonner,  1977).  Devine  (1966),  Phillips 
(1971),  and  McLachlin  and  Grindley  (1974)  note  the  importance  of  substrate 
stability  and  composition  in  limiting  distributions  of  burrowing  thalassinids.  How- 


600 


100 


200  300  400  500 

SODIUM    mM/LITER    MEDIUM 


FIGURE  4.  Blood  sodium  concentration  in  acclimated  C.  major  (solid  circles)  and  C. 
islagrande  (open  circles)  as  a  function  of  media  sodium  concentrations.  Each  large  solid 
circle  is  mean  of  6  to  9  determinations ;  each  small  solid  or  open  circle  is  value  for  individual 
animal;  vertical  lines  indicate  range.  Heavy  line  is  fitted  to  points  for  C.  major.  Asterisk 
denotes  juveniles. 


418 


DAKKYL  L.   FEI.DKK 


d 
O 

O 

_l 
m 

cr 

UJ 


10 


600 


500 


400 


300 


200 


100 


100  200  300  400  500 

SODIUM    mM/LITER    MEDIUM 


600 


FIGURE  5.  Blood  sodium  concentration  in  acclimated  C.  janniiccusc  as  a  function  of  media 
sodium  concentrations.  Each  solid  circle  is  mean  of  7  to  10  determinations;  vertical  lines 
indicate  range. 


ever,  both  substrate  and  salinity  are  thought  to  limit  penetration  of  Callianassa 
australicnsis  into  estuaries  (Hailstone  and  Stephenson,  1961). 

The  interaction  of  substrate  and  salinity  accounts  in  part  for  distributions 
of  Louisiana  Callianassidae ;  for  example,  C.  jamaiccnse  survives  at  high  salinities 
but  is  seldom  taken  above  25'/«,  because  substrates  in  those  areas  of  the  coast 
are  predominantly  sand  and  therefore  coarser  than  those  in  which  Phillips  (1971) 
reports  it  to  burrow  successfully.  Conversely,  C.  islagrandc  is  probably  limited 
to  transitory  occurrence  on  Grand  Terre  Island  and  ends  of  other  islands  by 
fluctuating  salinities,  since  sandy  substrates  in  those  areas  differ  little  from  sub- 
strates of  high  salinity  beaches  where  C.  islagrandc  is  common. 

Callianassa  major  and  adult  C.  islagrandc  cannot  osmoregulate  but  tolerate 
limited  reductions  of  salinity.  Similar  findings  are  reported  by  L.  Thompson  and 
Pritchard  ( 1969)  for  C.  calijomiensis  and  C.  filholi,  which  are  likewise  poikilos- 
motic  but  tolerate  salinities  down  to  —  10'/<  and  ~  13'/f,  respectively.  Osmo- 
conformation  and  limited  tolerance  of  dilute  media  are  also  reported  for  C. 
affinis  by  Gross  (1957).  It  thus  appears  that  the  polystenohaline  categorization, 
which  was  prematurely  applied  in  general  to  Callianassa  and  Upogcbia  by  earlier 
workers  (Lockwood,  1962;  Kinne,  1963),  may  be  retained  for  at  least  five  species 
of  Callianassidae  and  probably  for  others  which  occupy  similar  habitats.  However, 
some  of  these  species  are  less  stenohaline  than  others  ;  the  ability  of  Callianassa 
major  to  tolerate  acclimation  to  W/u  salinities  while  C.  islagrandc  usually  dies  at 


OSMOREGULATION  IX  CALLIANASSIDAE 


419 


this  salinity  in  part  explains  more  frequent  occurrence  of  C.  major  inside  the 
isohaline  and  its  predominance  on  ends  of  coastal  islands  where  salinities  occasion- 
ally fluctuate.  Preliminary  evidence  of  low-salinity  tolerance  and  slight  hyperosmotic 
regulation  in  juveniles  of  C.  islagrande  suggests  an  ontogenic  loss  of  tolerance 
and  regulatory  ability,  although  the  two  juveniles  studied  furnish  an  insufficient 
sample  for  firm  conclusions.  Juveniles  of  the  hermit  crab,  Pagnrus  bemhardus, 
regulate  volume  in  lower  salinities  than  adults,  and  Davenport  (1972a)  suggests 
that  the  aperture  of  the  nephropores  in  relation  to  body  size  limits  this  capacity 
in  adults. 

Tolerance  of  dilute  media  by  Callianassa  major  and  C.  islagrande  may  prove 
of  only  short-term  benefit  for  survival  of  populations.  Hill  (1971)  notes  that 
while  Upogebia  africana  can  tolerate  a  salinity  of  1.7'/,,  it  can  only  survive  through 
a  molt  in  a  salinity  >  3A(/l(.  Although  C.  major  and  adult  C.  islagrande  do  not 
appear  to  osmoregulate  (Fig.  1),  their  tolerance  of  dilute  media  may  in  part  relate 
to  accommodation  of  short-term  increases  in  blood  volume.  The  anterior  portion 
of  the  abdomen  is  soft,  and  its  elasticity  may  minimize  mechanical  effects  of  turgor. 
Davenport  (1972b)  suggests  such  an  adaptation  in  Pagnrus  bernhardns  and  shows 
that  with  increased  blood  volume  in  low  salinity,  a  larger  proportion  of  the  blood 
shifts  from  the  thorax  to  the  abdomen. 

The  degree  of  hypoionicity  in  blood  chloride  of  Callianassa   major  and   adult 


80 


Q    70 
O 

o 

m   60 
rr 

LL) 

t   50 


E  40 


ID 

c/>  30 

LU 


20 

10 


10        20         30        40        50        60         70 

MAGNESIUM     mM/LITER     MEDIUM 


80 


FIGURE  6.  Blood  magnesium  concentration  in  acclimated  C.  jamaicense  (solid  circles) 
and  C.  major  (open  circles)  as  a  function  of  media  magnesium  concentrations.  Each  large 
solid  or  open  circle  is  mean  of  6  to  10  determinations;  vertical  lines  indicate  standard  errors 
where  they  exceed  ±1.0;  small  solid  or  open  circles  are  individual  determinations. 


420 


DAKKYL   I..  FELDER 


+  2- 
+  1- 

0 
-1- 

_  2  — 

:E 

0     -3- 
tu 

^ 

LU 

<    -5H 


kw  I 

LU 

i    +3- 

X 

o+2- 

^    +1- 
n 

[           ( 

R                    1            "          rl 

u 

-1- 

n     x  7      T             1                 B 

-2- 

1 

-3- 


0.5  1.0  2.0  4.0 

DAYS,    EXPOSURE    TIME 

FIGURE  7.  Percentage  of  change  in  weight  of  body  water  at  timed  intervals  after 
direct  transfer  of  C.  jamaiccnsc  from  20/}-'f  salinity  to  3'/,(,  (open  circles)  or  37'/«,  (solid 
circles).  Each  solid  or  open  circle  is  mean  of  5  determinations;  rectangles  indicate  standard 
errors ;  vertical  lines  indicate  range. 

C.  islagrande  (Fig.  3)  is  very  near  that  reported  for  C.  californiensis  by  L.  Thomp- 
son and  Pritchard  (1969).  Some  degree  of  ionic  regulation  is  common  to  osmotically 
conforming  crustaceans,  but  levels  of  blood  chloride  in  such  crustaceans  are  usually 
reported  to  approximate  those  of  the  media  (Robertson,  1960;  Potts  and  Parry, 
1964).  L.  Thompson  and  Pritchard  (1969)  suggest  that  chloride  hypoionicity 
may  be  attributable  to  a  protein  anionic  component  of  blood  in  C.  coliforniensis. 
However,  as  noted  by  Dall  (1974),  blood  chloride  is  virtually  equivalent  to  blood 
sodium  at  any  given  salinity  despite  the  apparent  difference  when  blood  ion  con- 
centrations are  plotted  against  media  concentrations  of  the  same  ion.  Hence, 


OSMOREGULATION  IX  CALLIAXASSIDAK 


421 


where  sodium  and  chloride  in  media  are  at  normal  S\Y  ratios,  sodium  concen- 
tration heing  slightly  less  than  that  of  chloride,  equilihrium  between  the  two  ions 
is  reflected  in  hypoionicity  of  chloride  at  any  given  medium  concentration  of 
chloride  provided  blood  sodium  is  near  or  below  sodium  concentrations  of  the1 


00  - 


1000^ 


O      900 
O 

_) 
QQ 


^     800 

O 

2 
CO 

o 


700  - 


600  - 


500^ 


-37  %0     Medium 
1107.6  ±  4.3 


~3%o    Medium 
114.8  ±  0.3 


0.5 


.0 


2.0 


4.0 


9.0 


DAYS,    EXPOSURE    TIME 


FIGURE  8.  Blood  osmotic  concentration  at  timed  intervals  after  direct  transfer  of  accli- 
mated C.  jainaiccnsc  from  20'/f,  salinity  to  3'/rf  (open  circles)  or  to  37'/ef  (solid  circles). 
Each  solid  or  open  circle  is  mean  of  5  determinations ;  vertical  lines  indicate  range ;  rectangles 
indicate  standard  errors  where  they  exceed  ±10.  Figures  beneath  salinities  indicate  means 
and  standard  errors  of  media  osmotic  concentrations  over  9-day  period. 


422 


DARK VI.   L.   FKl.DKK 


~37%o    Medium 
Cl~  =  591.2  ±  1.2 
No"1"  =  492.9  ±  0.7 


0.5  1.0  2.0  4.0 

DAYS,     EXPOSURE      TIME 


9.0 


FIGURE  9.  Blood  chloride  (open  and  solid  circles)  and  sodium  (open  and  solid  squares) 
concentration  at  timed  intervals  after  direct  transfer  of  acclimated  C.  jnnniicciisc  from  20#r 
salinity  to  3'/,r  (open  circles  and  squares)  or  to  37'/,,  (solid  circles  and  squares).  Each 
circle  or  square  is  mean  of  5  determinations ;  vertical  lines  indicate  standard  errors  where 
they  exceed  ±5.  Figures  beneath  salinities  indicate  means  and  standard  errors  of  media 
ion  concentrations  in  mM/liter  over  9-day  period. 

medium,  l.lond  chloride  in  Callnnuissa  (Fig.  3)  exceeds  blood  sodium  concentra- 
tions (Figs.  4  and  5)  at  each  acclimation  salinity  and  the  degree  to  which  it  does 
so  increases  with  increasing  salinity,  probably  in  electrochemical  response  to  in- 
creased concentrations  of  magnesium  and  other  cations.  Blood  sodium  in  C. 


OSMOREGULATIOX  IX  CALLIANASSI1  ).\K 


423 


ma /or  and  adult  C.  islayrandc  is  equivalent  to  media  concentrations  <  300  HIM/ 
liter  and,  much  as  blood  osmolality  (Fig.  1),  drops  slightly  below  equilibrium  at 
the  upper  extremes  of  salinity.  Blood  sodium  and  osmotic  concentrations  respond 
similarly  in  acclimations  of  C.  calijorniensis,  but  both  sodium  and  osmolality  ol 


~37%0     Medium 
58.5  ±   0.3 


20  - 


0.5  1.0  2.0  4.0 

DAYS,    EXPOSURE    TIME 


9.0 


Fic.rkK  10.  Blood  magnesium  concentration  at  timed  intervals  after  direct  transfer  of 
acclimated  C.  jninaiccusc  from  20;,',  salinity  to  3'/f,  (open  circles)  or  to  37';,  (solid  circles). 
Each  solid  or  open  circle  is  mean  of  5  determinations;  vertical  lines  indicate  range; 
rectangles  indicate  standard  errors.  Figures  beneath  salinities  indicate  means  and  standard 
errors  of  media  concentrations  in  niM/liter  over  9-day  period. 


424  DARK VI.  L.  l-'KU)Kk 

blood  remain  more-  nearly  equivalent  to  media  concentrations  over  the  entire  range 
of  salinity  (L.  Thompson  and  Pritchard,  1969). 

Most  marine  crustaceans  strongly  hyporegulate  blood  magnesium  (Robertson, 
1953).  Exceptions  to  this  rule  include  several  brachyuran  spider  crabs,  the  primi- 
tive brachyuran,  l)roniia  vulgaris,  and  the  anomuran,  Lithodcs  inoia,  in  which  rela- 
tively high  blood  magnesium  is  correlated  with  low  levels  of  responsiveness 
attributed  to  magnesium  interference  with  neuromuscular  transmission  (Robertson, 
1960).  Callianassa  major  and  C.  jamaiccnse  also  have  high  levels  of  blood 
magnesium  and  appear  to  hyper-regulate  this  ion  at  media  concentrations  <  50 
mM/liter  (Fig.  6).  An  advantage  of  high  blood  magnesium  is  suggested  by  its 
effects  on  oxygen  binding  in  hemocyanins  (Larimer  and  Riggs,  1964;  Roxby, 
Miller,  Blair, "and  Van  Holde.  1974;  Miller  and  Van  Holde,  1974).  Miller  and 
Van  Holde  (1974)  report  a  mean  magnesium  concentration  of  48  mM/liter  for 
C.  californicnsis  at  an  unspecified  salinity.  Although  high  compared  to  that  of 
most  crustaceans,  this  value  is  well  within  the  ranges  of  blood  magnesium  here 
reported  for  C.  major  and  C.  jamaiccnse.  Specifically,  Miller  and  Van  Holde 
show  that  magnesium  effects  allosteric  transitions  in  callianassid  hemocyanin  in 
vitro.  By  increasing  oxygen  binding  (lowering  Pr,0)  high  blood  magnesium  may 
lie  advantageous  to  thalassinids  which  burrow  in  hypoxic  substrates.  Such  sub- 
strates are  inhabited  by  C.  californicnsis  on  the  Pacific  coast  (R.  Thompson  and 
Pritchard,  1969)  and  by  the  Callianassa  species  on  the  Louisiana  coast  (Felcler, 
in  preparation).  Miller  and  Van  Holde  (1974)  suggest  that  magnesium  levels 
remain  stable  in  Callianassa;  this  does  not  apply  to  C.  major  and  C.  jamaiccnse 
as  blood  magnesium,  while  somewhat  regulated,  varies  markedly  with  changing 
salinity.  Survival  of  these  species  in  a  hypoxic  habitat  could  thus  be  influenced 
by  interactions  between  salinity,  ion  balance,  and  oxygen  availability. 

Hyperosmotic  regulation  by  C.  jamaiccnse  at  low  salinities,  its  tolerance  of 
salinities  <  2(/((  and  its  ability  to  withstand  abrupt  changes  in  salinity  with  marked 
regulation  of  volume  clearly  support  its  categorization  as  euryhaline.  Such  capa- 
cities are  well  documented  among  upogebid  Thalassinidea  (Zenkevich,  1938;  L. 
Thompson  and  Pritchard,  1969;  Hill,  1971),  but  the  South  African  C.  kraussi 
is  the  only  other  species  of  the  Callianassidae  (sensit  de  Saint  Laurent,  1973) 
in  which  hypersomotic  regulation  is  reported  (Forbes,  1974).  The  blood  osmotic, 
sodium,  and  chloride  concentrations  in  acclimated  C.  jamaiccnse  (Figs.  2,  3, 
and  5)  closely  resemble  those  reported  for  C.  kraussi.  The  deterioration  of  regula- 
tory ability  that  Forbes  reported  in  C.  kraussi  at  lower  extremes  of  salinity  is  not 
pronounced  in  C.  jauiaicense,  probably  because  the  lowest  acclimation  extreme 
used  for  C.  kraussi  is  lower  than  that  used  for  C.  jauiaicense. 

The  difference  between  summer  and  winter  levels  of  hyperosmotic  regulation  in 
C.  jauiaicense  (Fig.  2)  likely  reflects  the  lower  field  temperatures  from  which 
winter  animals  were  collected;  both  field  temperature  and  salinity  were  lower 
during  winter  collections.  Lynch,  Webb,  and  Van  Engel  (1973)  and  Charmantier 
(1975)  list  a  number  of  studies  documenting  seasonal  temperature  effects  upon 
blood  osmotic  and  ionic  concentrations  in  crustaceans.  Higher  blood  osmotic 
and  ionic  concentrations  occur  in  animals  from  colder  water  (Dehnel,  1962; 
Ballard  and  Abbott,  1969),  even  when,  as  in  the  present  case,  acclimations  are 


OSMOREGULATION  IN  CALLIANASSIDAE  425 

conducted  at  equivalent  temperatures  in  the  laboratory.  Acclimation  studies  of 
Calimcctes  sapidns  suggest  that  lower  salinity  could  produce  an  effect  opposite 
from  that  of  low  temperature,  as  blood  osmotic  concentration  of  acclimated  crabs 
is  lower  when  crabs  are  collected  at  low  field  salinity;  however,  in  salinities  <  15'/,. 
blood  osmotic  concentrations  of  Callinectcs  sapidits  depend  little  upon  the  direction 
from  which  the  acclimation  salinity  is  approached  (Ballard  and  Abbott,  1969). 

After  direct  transfers  of  Callianassa  jauiaicensc  from  20'/r,  the  animals  placed 
into  3'/f  regulate  volume  nearer  original  levels  than  do  those  placed  into  37%o, 
but  in  both  cases  volumes  are  near  original  levels  after  two  clays  (Fig.  7). 
Limited  data  on  weight  changes  of  C.  major  and  C.  islayrandc  after  less  dramatic 
stepwise  transfers  to  low  salinities  suggest  much  poorer  volume  control  in  those 
species  (Table  II).  The  means  by  which  C.  jauiaicensc  controls  volume  and 
blood  osmotic  concentration  is  at  present  uncertain.  Studies  of  urine  in  both 
hyperosmotically  regulating  (Forbes,  1974)  and  osmotically  conforming  (L. 
Thompson  and  Pritchard,  1969)  Callianassidae  show  that,  as  in  the  great  majority 
of  euryhaline  Crustacea  (Potts  and  Parry,  1964),  an  isosmotic  urine  is  produced 
by  animals  once  acclimated  to  various  salinities.  However,  urine  volumes  and 
osmolality  are  not  reported  during  acclimation  in  either  of  these  studies.  Osmo- 
regulatory  functions  of  the  antennal  glands  are  suggested  by  increased  urine 
volumes  in  the  crab,  Carcinns  inaenas,  with  decreased  salinity  (Binns,  1969)  and 
by  studies  of  the  lobster,  Homanis  aincricaniis,  in  which  urine  is  near  isosmotic  to 
blood  in  animals  fully  acclimated  to  lowered  salinity,  but  markedly  hyposmotic 
during  acclimation  (Dall,  1970).  Changes  in  permeability  may  also  facilitate 
regulation  of  volume  and  blood  osmolality,  and  such  changes  are  documented  in 
other  euryhaline  decapods  subjected  to  dilute  media  (Capen,  1972;  Spaargaren, 
1975).  Additionally,  Heeg  and  Cannone  (1966)  describe  an  osmoregulatory 
diverticulum  on  the  posterior  mid-gut  of  grapsid  crabs;  a  similar  diverticulum  is 
present  in  Callianassa  jainaicense,  C.  major,  and  C.  islat/rande,  although  its  function 
is  unknown. 

After  direct  transfer  of  C.  jai/iaiccnse  from  2Q'/,  to  3'/<(,  media,  blood  osmotic, 
chloride,  and  sodium  concentrations  are  near  new  stable  levels  within  12  hours, 
but  gradual  changes  in  the  sodium/chloride  ratio  continue  to  occur  through  day  9 
(Figs.  8  and  9).  Changes  in  blood  osmotic  and  sodium  concentrations  of  C. 
jauiaicensc  are  very  nearly  proportional  over  observed  time  increments  after 
transfers  to  either  3'/,<  or  37  %c  media.  A  similar  close  correlation  between  sodium 
and  osmotic  concentrations  is  reported  in  crustacean  blood  by  other  investigators 
(Colvocoresses,  Lynch,  and  Webb,  1974)  and  such  observations  seem  compatible 
with  data  indicating  that  the  sodium  transport  system  ultimately  establishes  the 
blood  osmolality  (Shaw,  1960).  By  day  9  after  direct  transfers,  blood  sodium/ 
chloride  ratios  in  C.  jauiaicensc  are  higher  at  3(/<f  than  at  37r/c  salinity.  A  similarly 
elevated  sodium/chloride  ratio  is  also  observed  after  C.  kranssi  is  acclimated  to 
low  salinity  (Forbes,  1974). 

Blood  magnesium  concentrations  approach  stable  levels  less  rapidly  than  other 
ions  after  salinity  transfers  (Fig.  10).  This  may  contribute  to  what  Forbes 
(1974)  describes  as  slower,  smaller  changes  in  blood  osmotic  concentrations  after 
stabilization  of  blood  sodium  and  chloride  concentrations  following  salinity 
transfers  of  C.  kranssi;  Forbes  (p.  310)  speculates  such  changes  could  be  asso- 


426  I). \RRV1.   L.   FKI.DHK 

ciated  with  "non-ionic  osmotically  active-  entities  in  the  blood,"  lint  dors  not  report 
divalent   ion   concentrations. 

Evolution  of  hyperosmotic  regulation  in  C.  kranssi  is  attributed  to  the  unique 
flood-influenced  salinity  gradient  in  southern  African  estuaries  (Forbes,  1974)  ; 
similar  conditions  occur  in  other  areas  including  coastal  estuaries  of  the  Northern 
Gulf  of  Mexico  (Hewatt,  1951  ;  Barrett,  Tarver,  Latapie,  Pollard,  Mock,  Adkins, 
Gaiclry,  White,  and  Mathis,  1971).  Euryhalinity  may  be  characteristic  of  a  phyletic 
stock,  rather  than  of  an  isolated  species  or  genus  and  probably  is  a  very  conservative 
physiological  adaptation  once  acquired  (Hedgpeth,  1957);  Ortmann  (1902) 
furnishes  examples  of  such  phyletic  stocks  among  crustaceans  for  the  now  fresh- 
water Atyidae  and  the  Palaemonidae  which  occur  in  marine,  estuarine,  and 
freshwater  habitats.  Since  Callianassa  kraussi  and  C.  jamaicense  share  the  con- 
servative character  of  euryhalinity,  further  examination  of  their  phylogenetic 
proximity  may  prove  interesting.  However,  phylogenetic  interpretations  must 
be  made  with  caution;  osrnoregulatory  ability  may  be  a  conservative  trait  once 
acquired,  but  it  could  have  been  acquired  independently  following  separation  of 
ancestral  stocks.  Lockwood  and  Croghan  (1957)  suggest  that  only  700  years 
were  required  for  development  of  a  separate  race  of  Baltic  isopods  which  now 
possesses  distinctly  greater  powers  of  osmotic  and  ionic  regulation  than  its 
ancestral  stocks. 


I  wish  to  thank  Dr.  J.  Porter  Woodring  for  his  helpful  advice  and  criticisms 
during  both  the  research  and  writing  phases  of  this  study.  I  also  thank  S.  Felder 
and  K.  Vincent  who  assisted  in  typing  and  proofing  the  manuscript. 

SUMMARY 

Osmotic  and  ionic  regulatory  capacities  of  callianassid  mud  shrimps,  Callianassa 
jainaicense,  C.  major,  and  C.  islagrande,  are  correlated  to  their  distributions  on 
the  Louisiana  coast.  Callianassa  jamaicense  burrows  in  muddy  estuaries  where 
salinity  may  commonly  fall  to  <  Sl/<(.,  but  C.  major  and  C.  islagrandc  usually  burrow 
in  sandy  beaches  bathed  by  higher  salinities.  Lower  lethal  limits  of  salinity  are 
<  2l/,(!  for  C.  jainaicense,  7-S'/(  for  C.  major  and  probably  just  below  I5f/(c  for 
adult  C.  islayrande.  After  exposure  to  low  salinity  C.  jainaicense  exhibits  better 
volume  control  than  the  other  two  species.  Blood  osmotic,  sodium,  and  chloride 
concentrations  in  C.  jamaicense  are  regulated  near  stable  levels  at  acclimation 
salinities  beneath  <— '  2Q'/«  but  those  of  C.  major  and  C.  islagrandc  are  not.  Blood 
magnesium  is  slightly  hyper-regulated  by  C.  jamaicense  at  most  acclimation  salin- 
ities <  25'/,{  and  more  markedly  hyper-regulated  at  salinities  <  10'/<  ;  it  is  also 
slightly  hyper-regulated  by  C.  major  at  acclimation  salinities  <  30'/r. 

After  direct  transfer  of  C.  jamaicense  from  20'/r  salinity  to  3'/'<  salinity,  blood 
osmotic,  sodium,  and  chloride  concentrations  fall  slightly  but  approach  stable  con- 
centrations within  12  hours;  blood  magnesium  concentration  falls  less  rapidly. 
When  C.  jamaicense  is  transferred  from  20  to  37^,,  blood  osmotic,  sodium,  and 
chloride  concentrations  increase  markedly  during  the  first  day  and  continue  to 


OSMOREGULATION  IN  CALLIANASSIDAE  427 

slowly  increase  through  day  9;  blood  magnesium  increases  to  a  near  stable  level 
by  day  4. 

Differences  in  osmoregulatory  capacities,  along  with  substrate  preferences, 
appear  to  limit  distributions  of  Callianassidae  on  the  Louisiana  coast.  With  one 
exception,  previous  studies  suggest  that  osmoregulatory  ability  does  not  occur  in 
this  group.  The  present  report  of  osmoregulatory  ability  in  C.  jaiimiccnsc  docu- 
ments a  second  exception. 

LITERATURE  CITED 

BALLARD,  B.  S.,  AND  W.  ABBOTT,  1909.     Osmotic  accommodation  in  Callinectes  sapid  us  Rathliun. 

Comp.  Biochcm.  Physiol.,  29A  :   671-687. 
BARNES,   R.   S.   K.,   1967.     The  osmotic  behavior  of  a  number  of  grapsoid  crabs   with   respect 

to  their  differential  penetration  of  an  estuarine  system.     /.  E.rp.  Biul.,  47:  535-551. 
BARRETT,  B.  B.,  J.  W.  TARVER,  W.  R.  LATAPIE,  J.   F.   POLLARD,   W.   R.   MOCK,  G.   B.  ADKINS, 

W.  J.  GAIDRY,  C.  J.  WHITE,  AND  J.  S.  MATHIS,   1971.     Phase  II,  Hydrology.     Pages 

9-130  in  B.  B.  Barrett,  Ed.,  Cooperative  Gitlf  of  Mexico  estimrine  inventory  and  study, 

Louisiana.     Louisiana  Wildlife  and  Fisheries  Commission,  New  Orelans. 
BINNS,  R.,  1969.     The  physiology  of  the  antennal  gland  of  Curcinns  macmis    (L. )    II.     Urine 

production  rates.    /.  Ex  p.  Biol.,  51  :  11-16. 
BROWN,   F.,   AND   W.   D.   STEIN,    1960.        Balance   of   water,   electrolytes,   and   non-electrolytes. 

Pages  403-470  in  M.   Florkin  and  H.  S.  Mason,   Eds.,  Comparative  biochemistry,   I'ul. 

II.    Academic  Press,  New  York. 
BURSEY,   C.   R.,   AND   E.   E.   BONNER,   1977.     Osmotic   regulation   and   salinity   tolerance   of   the 

mole  crab,  Emcrita  talpoida   (Say)    (Crustacea,  Anomura).     Comp.  Biochcm.  Physiol., 

57A:  207-210. 
BYBEE,  J.  R.,  1969.     Effects  of  hydraulic  pumping  operations  on  the  fauna  of  Tijuana  Slough. 

Calif.  Fish.   Game,  55:   213-220. 
CAPEN,  R.  L.,  1972.     Studies  of  water  uptake  in  the  euryhaline  crab,  Rhithropanopetts  liarrisi. 

J.  E.rp.  Zoo/.,  182:  307-319. 
CHABRECK,  R.   H.     1972.     Vegetation,   water  and   soil  characteristics  of  the   Louisiana  coastal 

region.    La.  State  Univ.  Auri.  Exp.  Sta.  Bull.,  664:  1-72. 
CHARMANTIER,  G.,  1975.     Variations  saisonnieres  des  capacities  iono-regulatrices  de  Spliacroma 

scrratum     ( Fabricius,     1787)      (Crustacea,     Isopoda,    Flabellifera) .      Comp.    Biochcm. 

Physiol.,  50 A  :  339-346. 
COLVOCORESSES,  J.  A.,  M.  P.  LYNCH,  AND  K.  L.  WEBB,  1974.     Variations  in  serum  constituents 

of  the  blue  crab,  Callinectes  sapidns:  Major  cations.     Comp.  Biochcm.  Physiol.,  49A  : 

787-803. 
DALL,  W.,  1970.     Osmoregulation  in  the  lobster  Homarus  americanus.     J.  Fish.  Res.  Bd.  dm., 

27:  1123-1130. 
DALL,  W.,  1974.     Osmotic  and  ionic  regulation  in  the  western  rock  lobster     Pitniilints  loin/ipcs 

(Milne-Edwards).     /.  E.rp.  Mar.   Biol.  Ecol..   15:   97-125. 
DAVENPORT,  J.,    1972a.     Effects   of   size   upon   salinity  tolerance   and   volume   regulation   in  the 

hermit  crab  Paynnts  bcrnhardus.     Mar.  Biol.,  17  :  222-227. 
DAVENPORT,  J.,  1972b.     Study  of  the  importance  of  the  soft  abdomen  of  the  hermit  crab  Pat/urns 

hcrnhanhts   in   minimizing   the   mechanical   effects   of  osmotic    uptake   of   water.      Mar. 

Biol.,   17:  304-307. 
DEHNEL,  P.  A.,  1962.    Aspects  of  osmoregulation  in  two  species  of  intertidal  crabs.     Biol.  Bull., 

122:  208-227. 
DEVINE,  C.  E.,   1966.     Ecology  of  Callianassa  filholi  Milne-Edwards   1878    (Crustacea,  Thalas- 

sinidea).     Trans.  Roy.  Soc.  N.  Z.,  8:  93-110. 
DEWINDT,  J.  T.,  1974.     Callianassid  burrows  as  indicators  of  subsurface  beach  trend,  Mississippi 

River  Delta  Plain.    J.  Sediment.  Petrol.,  44:  1136-1139. 
FORBES,  A.  T.,  1974.     Osmotic  and  ionic  regulation  in  Callianassa  kraiissi  Stebbing   (Crustacea: 

Decapoda:  Thalassinidea).     /.  E.rp.  Mar.  Biol.  Ecol..   16:  301-311. 


428  DARRYL  L.  FELDER 

FRANKKNBERG,  D.,  S.  L.  COLES,  AND  R.  E.  JOHANNES,  1967.     The  potential  trophic  significance 

of  Callianassa  major  fecal  pellets.     Linnwl.  Oceanogr.,   11:    191-197. 
GKOSS,  W.,  1957.     An  analysis  of  response  to  osmotic  stress  in  selected  Crustacea.     /Hal.   Hull., 

112:  43-62. 
HAILSTONE,  T.  S.,  AND  W.  STEPHENSON,  1961.    The  biology  of  Callianassa  (Try pen)  australicn- 

sis  Dana  1852  (Crustacea,  Thalassinidea).    Univ.  Quccnsl.  Pap.  Dcp.  Zoo!.,  1:  259-285. 
HEDC.rF.TH,    J.    W.,    1950.      Notes    on    the    marine    invertebrate    fauna    of    salt    flat    areas    in 

Aransas    National    Wildlife    Refuge,    Texas.      Pnbl.    Inst.    Mar.    Sci.    Univ.    Tex.,    1  : 

103-119. 
HEDGPETH,   J.   W.,   1957.      Estuaries  and  lagoons:    II.      Biological   aspects.      Pages   693-729   in 

J.   W.   Hedgpeth,   Ed.,   Memoir  67,   Treatise  on   marine   ecology   and   paleoecology,    1. 

Ecoloi/v.     Geological   Society  of  America,   New   York. 
HEEG,   J.,   AND   A.   J.   CANNONE,    1966.      Osmoregulation   by   means   of   a   hitherto   unsuspected 

osmoregulatory  organ  in  two  grapsid  crabs.     Zool.  Afr.,   11:   127-129. 
HEWATT,    W.    G.,     1951.      Salinity    studies    in    Louisiana    coastal    cmbaymcnts    west    of    the 

Mississippi  River:  final  report  of  Project  Nine.     Texas  A  &  M  Research  Foundation, 

College   Station,  32  pp. 
HILL,  B.  J.,  1971.    Osmoregulation  by  an  estuarine  and  a  marine  species  of  Upogcbia  (Anomura, 

Crustacea).    Zool.  Afr..  6  :  229-236. 
KINNE,    O.,    1963.     Adaptation,   a   primary   mechanism   of   evolution.      Pages   27-50   in    H.    B. 

Whittington  and  W.  D.  I.  Rolfe,  Eds.,  Phylogcny  and  evolution  of  Crustacea.     Special 

Publication  of  the  Museum  of  Comparative  Zoology,   Cambridge. 
LARIMER,  J.  L.,  AND  A.  F.  RIGGS,  1964.     Properties  of  hemocyanins — I.     The  effect  of  calcium 

ions  on  the  oxygen  equilibrium  of  crayfish  hemocyanin.     Coinp.  Biochem.  Physiol.,  13  : 

35-46. 
LELOEUFF,   P.,  AND  A.   INTES,    1974.     Les   Thalassinidea    (Crustacea,   Decapoda)    du   Golfe   de 

Guinee,  systematique — ecology.     Call.  Office  dc  la  Recherche  Scientifique  ct  Technique 

Oittrc-Mcr  Ser.  Oceanogr.,  12:  17-69. 

LOCKWOOD,  A.  P.   M.,   1962.     The  Osmoregulation  of  Crustacea.     Biol.  Rev.,  37  :   257-305. 
LOCKWOOD,  A.  P.  M.,  AND  P.  C.  CROGHAN,  1957.     The  chloride  regulation  of  the  brackish  and 

fresh-water  races  of  Mcsidotca  cntomon    ( L. ) .     /.  E.vp.  Biol.,  34:  253-258. 
LUNZ,  G.  R.,  1937.     Notes  on  Callianassa  major  Say.     Charleston  Museum  Leaflet,  10:   1-15. 
LYNCH,  M.  P.,  K.  L.  WEBB,  AND  W.  A.  VAN   ENGEL,   1973.     Variation  in  serum  constituents 

of   the    blue    crab,    Callinectes    sapid  us:    chloride    and    osmotic    concentration.      Com  p. 

Biochem.  Physiol.,  44A  :  719-734. 
McLACHLiN,    A.,    AND   J.    R.    GfiiNDLEY,    1974.      Distribution    of    rnacrobenthic    fauna    of    soft 

substrata  in  Swartkops  Estuary.     Zool.  Afr.,  9:  211-233. 
MILLER,    K.,    AND    K.    E.    VAN    HOLDE,    1974.      Oxygen    binding    by    Callianassa    californiensis 

hemocyanin.     Biochemistry,   13  :    1668-1674. 
MONOD,   T.,   1927.     Sur  le  crustace  auquel   le  cameroun  doit   son   nom    (Callianassa   turnerana 

White).     Bull.  Mus.  Hist.  Nat.  Paris.  33:  80-85. 
ORTMANN,  A.  E.,  1902.     The  geographical  distribution  of  freshwater  decapods  and  its  bearing 

upon  ancient  zoogeography.     Proc.  Amer.  Philos.  Soc.,  41  :  267-400. 
PHILLIPS,    P.    J.,    1971.      Observations    on   biology    of    mudshrimps    of    the    genus    Callianassa 

(Anomura  :  Thalassinidea)    in  Mississippi  Sound.     Gulf  Res.  Rep.,  3  :   165-196. 
POHL,    M.   E.,    1946.      Ecological    observations   on    Callianassa    major    Say   at    Beaufort,    North 

Carolina.     Ecology,  27:   71-80. 
POTTS,  W.  T.  W.,  AND  G.  PARRY,  1964.     Osmotic  and  ionic  regulation  in  animals.     Pergamon 

Press,  Oxford,  423  pp. 
ROBERTSON,  J.  D.,  1953.     Further  studies  on  ionic  regulation  in  marine  invertebrates.     /.  Exp. 

Biol.,  30:  277-296. 
ROBERTSON,  J.  D.,  1960.     Osmotic  and  ionic  regulation.     Pages  317-339  in  T.   H.   Waterman, 

Ed.,  Physiology  of  Crustacea.     Academic   Press,   New   York. 
RODRIGUES,  S.  A.,  1971.     Mud  shrimps  of  the  genus  Callianassa  Leach  from  the  Brazilian  coast 

(Crustacea,   Decapoda).     Arq.   Zool.    (Sao   Paulo),    20:    191-223. 
ROXBY,  R.,  K.  MILLER,  D.  P.  BLAIR,  AND  K.  E.  VAN   HOLDE,   1974.     Subunits  and  association 

equilibria  of  Callianassa  hemocyanin.     Biochemistry,  13  :   1662-1668. 
SAINT  LAURENT,  M.  DE,  1973.     Sur  la  systematique  et  la  phylogenie  des  Thalassinidea  :  defini- 


OSMOREGULATION  IX  CALLIAXASSIDA  I  429 

tion  des  families  des  Callianassidae  et  cles  Upogcbiidae  et  diagnose  de  cinq  genres  nou- 
veaux  (Crustacea  Decapoda).  C.  R.  Acad.  Sci.  Paris,  277:  513-516. 

SHAW,  J.,  1960.  The  absorption  of  chloride  ions  by  the  crayfish  Astacus  paUipcs.  J.  E.vp. 
BioL.  37  :  557-572. 

SNELLING,  B.,  1959.  The  distribution  of  intertidal  crabs  in  the  Brisbane  River.  Aiist.  J.  Mar. 
Frcsln^itcr  Res.,  10  :  67-81. 

SPAARGAREN,  D.  H.,  1975.  Energy  relations  in  the  ion  regulation  in  three  crustacean  species. 
Comp.  Biochcm.  Physio!.,  51A  :  543-548. 

TEAL,  J.  M.,  1958.     Distribution  of  fiddler  crabs  in  Georgia  salt  marshes.     Ecology,  39  :  185-193. 

THOMPSON,  L.  C.,  AND  A.  W.  PRITCHARD,  1969.  Osmoregulatory  capacities  of  Callianassa 
and  Upoycbia  (Crustacea:  Thalassinidea).  Biol.  Bull.,  136:  114-129. 

THOMPSON,  R.  K.,  AND  A.  W.  PRITCHARD,  1969.  Respiratory  adaptations  of  two  burrowing 
crustaceans,  Culliaiiassa  califoniicusis  and  Upoycbia  pnycttcnsis  (Decapoda,  Thalas- 
sinidea). Biol.  Bull..  136  :  274-287. 

WASS,  M.  L.,  1955.  The  decapod  crustaceans  of  Alligator  Harbor  and  adjacent  inshore  areas 
of  northwest  Florida.  Q.  J.  Fla.  Acad.  Sci.,  26:  1-179. 

WEIMER,  R.  J.,  AND  J.  H.  HOYT,  1964.  Burrows  of  Callianassa  major  Say  as  indicators  of 
littoral  and  shallow  neritic  environments.  /.  PaleontoL,  38  :  761-767. 

WILI.IS,  E.  R.,  1942.  Some  mud  shrimps  of  the  Louisiana  coast.  Occas.  Papers  Afar.  Lab. 
La.  State  Univ.,  2  :  1-6. 

ZENKEVICH,  L.  A.,  1938.  The  influence  of  Caspian  and  Black  Sea  waters  of  different  concen- 
tration upon  some  common  Black  Sea  invertebrates.  Part  II.  The  change  in 
internal  salinity.  Zool.  Zh..  17:  976-1002  (translation  from  Russian). 


Hinl.  Hull.,  154:  430-439.  Hum-.  W7X) 


LARVAL  REARING,  METAMORPHOSIS,  GROWTH  AND  REPRODUC- 
TION  OF  THE   EOLID   NUDIBRANCH   HERMISSENDA    CRASSI- 
CORNIS   (ESCHSCHOLTZ,   1831)  (GASTROPODA: 
OPISTHOBRANCHIA) 

JUNE  F.  HARRIGAN  AND  DANIEL  L.  ALKON 

Laboratory    i>f    Hiophysics,    Section    on    Ncunil    Systems,    Intramural    Research    Prouram, 

National    Institute    of    Neurological    and    Communicative    Disorders    and    Stroke, 

National  Institutes   of   Health.   Marine   Biological  Laboratory, 

Woods   Hole,   Massachusetts  02543 

In  the  eolid  nudibranch,  Hermissenda  crassicomis  (Eschschultz,  1831),  neural 
pathways  responsive  to  light,  chemosensory  stimuli,  and  gravitational  stimuli  con- 
verge within  the  circumesophageal  nervous  system  (Alkon,  1974,  1975,  1976). 
These  convergence  points,  as  defined  by  intracellular  recordings,  may  he  important 
for  choice  behavior  and  behavioral  modification  as  demonstrated  for  this  animal. 
Maintenance  conditions,  primarily  light-dark  cycle,  temperature,  and  diet,  had  to 
he  carefully  controlled  when  analyzing  both  the  behavior  and  the  nervous  system  of 
Hermissenda.  The  goal  of  the  present  study  was  to  establish  a  laboratory  strain  of 
Hermissenda  to  provide  animals  of  known  history  for  these  studies,  and  for  studies 
on  behavioral  and  neural  development  in  these  three  sensory  pathways. 

Hermissenda,  a  monotypic  genus,  is  widely  distributed  along  the  west  coast 
of  North  America  (Lance,  1966;  MacFarland,  1966).  Field  observations  (Yarnell, 
1972;  Birkeland,  1974)  indicate  that  Hermissenda,  although  preying  primarily  on 
coelenterates,  has  a  broader  diet  than  most  nudibranchs.  Hermissenda  and  its  egg 
masses  appeared  on  fouling  panels  exposed  for  one  month  at  a  time  throughout 
the  year  in  Monterey  Bay,  California  (Haderlie,  1968).  Year-round  availability 
of  eggs  and  adults  and  a  relatively  broad  diet  in  the  adult  stage  simplify  cultivation 
of  Hermissenda. 

MATERIALS  AND  METHODS 
Reproductive  periodicity 

Weekly  shipments  of  Hermissenda  were  obtained  from  Mr.  Michael  Morris, 
Sea  Life  Supply,  Sand  City,  California.  Ten  animals  (2.5  +  cm  in  length)  were 
removed  from  each  week's  shipment  from  May,  1976,  to  May,  1977.  Animals 
were  incubated  separately  on  a  12L:  12D  schedule  at  an  average  sea  water  tem- 
perature (±s.d.)  of  14.1°  ±  1.8°  C,  which  approximates  the  mid-point  of  the 
annual  temperature  range  occurring  in  the  natural  habitat  (9°-18°  C;  Haderlie, 
Mellor,  Minter,  and  fteoth,  1975).  A  daily  record  was  kept  of  the  number  of  each 
set  of  ten  animals  that  deposited  an  egg  mass. 

Fecundity   measurements 

Fifty  newly-arrived  animals  of  widely  varying  sizes  (73-3204  mg)  were 
weighed  underwater  on  a  Mettler  PN323  balance,  immediately  after  each  deposited 

430 


CUI/riVATlON  OF  HKRMISSKXD.I  431 

its  first  egg  mass.    The  number  of  eggs  per  egg  mass  was  estimated  l>y  multiplying 
the  length  of  the  egg  string  in  nun  by  the  average  number  of  embryos  per  mm. 

Egg  diameter  and  egg  capsule  size  were  measured  through  a  Zeiss  Universal 
microscope  with  a  calibrated  ocular  micrometer.  All  computer  analyses  of  data  on 
growth  and  reproduction  were  performed  on  a  PDF  1 1/10  computer,  using  standard 
statistical  packages. 

Hief   experiments 

Forty  individuals  (ten  per  diet)  were  maintained  until  death  on  one  of  four 
locally  available  diets:  frozen  squid  mantle  muscle  ( Lolit/o  pealii),  mussel 
(Mytilus  ecinlis),  tunicate  (Ciona  intestinalis),  minus  the  tunic,  and  an  alternating 
diet  consisting  of  one  day  of  squid,  then  mussel,  then  tunicate,  etc. 

Weighed  animals  were  placed  singly  in  numbered  7x4  cm  plastic  snap-top 
vials  which  were  perforated  with  slits  for  water  exchange,  and  maintained  at  a  sea 
water  temperature  of  12:-14°  C.  An  excess  of  food  was  provided  fresh  daily. 
Weight,  days  survived,  number  of  egg  masses  laid  and  whether  eggs  developed 
normally  were  recorded  for  each  animal  on  each  diet. 


Larval  rearing 

Egg  masses  were  incubated  at  13°-15°  C  in  0.22  ^m  Millipore-filtered  (MPF) 
sea  water.  On  day  5  or  6  following  oviposition,  the  veligers  were  liberated  by 
teasing  apart  the  egg  mass.  Cetyl  alcohol  flakes  sprinkled  on  the  surface  of  larval 
cultures  prevented  larvae  from  becoming  entrapped  in  the  surface  film  (Hurst, 
1967).  The  rearing  method  was  adapted  from  that  developed  for  aplysiid  larvae 
by  Switzer-Dunlap  and  Hadfield  (1978).  A  similar  method  was  employed  by 
Harrigan  and  Alkon  (1978)  to  rear  the  opisthobranch  molluscs,  Elysia  chlorotica 
Gould,  1870  and  Hauiinoea  solitaria  (Say,  1822). 

Larvae  were  cultured  at  a  concentration  of  three  per  ml  in  covered  one-liter 
Pyrex  beakers  filled  with  800  ml  of  culture  water  (0.22  /Jin  MPF  sea  water 
containing  5  ppm  chloromphenicol).  Cultures  were  maintained  on  a  12L:12D 
cycle  at  an  average  temperature  of  13.8°  ±  1.2°  C.  Larvae  were  transferred  three 
times  a  week  to  clean  culture  water.  Cultures  were  fed  daily  and  stirred  to 
resuspend  food  and  veligers. 

Cultures  were  initially  fed  equal  amounts  of  Isochrysis  t/alhana  and  Monochrysis 
Intheri  at  a  final  concentration  of  3.0  X  104  cells  per  ml  and  the  larger  flagellate 
Chroococcus  salina  (strain  3C  )  at  a  final  concentration  of  7.5  X  10*  cells  per  ml. 
Food  concentration  was  gradually  decreased  as  the  cultures  aged.  Algal  cultures 
were  bacteria-free,  and  were  grown  in  100  ml  aliquots  according  to  the  methods 
of  dullard  (1975). 

Juveniles  were  fed  for  one  week  on  the  hydroid  species  (provided  by  Sea  Life 
Supply)  on  which  they  metamorphosed.  They  were  then  fed  only  on  tunicate 
(dona  intestinalis)  obtained  from  Cape  Cod  waters.  Body  lengths,  measured 
when  the  animals  were  fully  extended,  were  taken  weekly;  body  weights  were 
taken  occasionally. 


432  j.  i;.  IIAKKK;.\\  AND  n.  L.  ALKON 

RESULTS 

Reproductive  periodicity 

Fertile  egg  masses,  which  produced  normal  veligers,  were  obtained  every  week 
of  the  year  from  sets  of  ten  animals  collected  from  Monterey  Bay,  California. 
Chi-square  analysis  of  an  R  X  C  contingency  table  indicated  no  significant  inter- 
action between  the  number  of  animals  laying  eggs  per  week  and  the  month  of 
the  year  the  eggs  were  obtained  (P  >  0.99,  df  ==  33).  Hermissenda  did  not  ex- 
hibit seasonal  periodicity  in  egg-laying  in  the  laboratory. 

Over  the  one-year  sampling  period  79%  of  the  total  number  of  animals  tested 
(n  =  490)  deposited  at  least  one  egg  mass.  Thirty-one  per  cent  of  the  animals 
laying  one  egg  mass  produced  a  total  of  2,  3,  or  4  egg  masses  within  the  one-week 
test  period. 

Characteristics  of  the  egg  mass 

Hermissenda  deposits  its  egg  masses,  while  or  pink,  in  a  tight  counter-clockwise 
spiral.  Structure  of  the  egg  mass  is  further  described  by  Hurst  (1967).  Diameters 
of  the  first  egg  mass  deposited  in  the  laboratory  by  the  adults  (73-3204  mg  body 
weight)  ranged  from  0.24  cm  to  3.62  cm.  Average  egg  mass  diameter  increased 
linearly  with  adult  weight  (polynomial  regression,  P  <  0.01,  df  =  49).  The 
number  of  eggs  estimated  per  egg  mass  (see  Methods)  varied  from  6.9  X  103  to 
1.0  X  10fi. 

The  number  of  eggs  per  egg  capsule  increased  with  adult  weight  (P  <  0.01) 
from  one  to  an  average  of  nine  eggs  for  adults  greater  than  500  mg.  Eggs  are 
packed  one  per  capsule  for  adults  weighing  less  than  500  mg.  The  average  egg 
diameter  was  65.4  ±  1.2  /zin  (n  =  70,  7  adults).  Egg  diameter  was  not  a  function 
of  the  number  of  eggs  per  capsule.  Egg  capsule  length  increased  significantly  with 
number  of  eggs  per  capsule  (one-way  ANVAR,  P  <  0.01)  (Table  I). 

Larval  development 

Veligers  hatch  in  5-6  days  at  13°-15°  C.  Unsculptured  shells  are  of  about 
f  whorl  and  belong  to  Thompson's  Type  I  (Thompson,  1961).  Average  shell 
length  and  width  at  hatching  is  105.9  ±  6.3  X  75.4  ±  4.8  Mm  (n  =  25).  Hermis- 
senda has  an  obligatory  veliger  stage  of  at  least  34  days.  Metamorphosis  is  delayed 

TABLE  I 

Relationship  between  capsule  size  and  number  of  eggs  per  capsule.     Each  number  represents  average 
length  and  width  (fim)  of  100  capsules,  20  from  each  of  five  adults. 

Eggs/capsule  Length  X  width  (^m)  (=ts.d.) 

1  110.3  ±  10.0  X  76.2  ±    4.4 

2  145.5  ±    5.8  X  102.3  ±    6.1 

3  157.0  ±    9.9  X  112.8  ±    3.8 

4  175.4  ±    8.5  X  126.4  ±    7.1 

5  185.8  ±  10.1  X  141.9  ±  10.9 


CULTIVATION  OF  HERMISSENDA 


433 


350  r 


300  - 


50 


0 


10 


20 
DAYS 


30 


40 


FIGURE  1.  Growth  of  veligers  in  terms  of  shell  length.  Vertical  bars  represent  one 
standard  deviation.  Dots  represent  average  shell  length ;  triangles  represent  the  size  of  the 
largest  individual  measured. 

by  about  2-4  clays  after  maximum  sbell  length  (310.4  ±  9.8  /xin,  n  =:  11 )  is  attained. 

Veligers  which  were  competent  to  metamorphose  were  recognized  by  the 
following  criteria :  presence  of  eyes,  shell  length  of  at  least  300  p,m,  enlargement 
of  the  foot  and  development  of  the  propodium,  reduced  swimming  activity,  the 
veliger  remaining  on  or  near  the  bottom,  and  the  presence  of  a  tooth  at  the  base 
of  the  shell  aperture.  The  average  shell  length  of  a  sample  of  veligers  did  not 
accurately  reflect  the  size  of  the  largest  individuals.  Figure  1  illustrates  shell 
growth  in  veligers  from  three  replicate  cultures. 

On  day  34  post-hatching,  competent  veligers  crawled  immediately  on  the  thecate 
hydroid,  Obclia  longissiina,  and  on  an  unidentified  thecate  hydroid  from  Cali- 
fornia. Competent  veligers  also  crawled  on  the  related  species,  Obclia  ijcnicnlata, 
collected  from  Woods  Hole,  Massachusetts.  The  velum  is  lost  during  the  first 
12-24  hours  after  crawling  begins.  In  the  next  12-24  hours  the  larva  slowly 
crawls  out  of  its  shell.  During  shell  exit  one  pair  of  tentacle  buds  and  two  pairs 
of  cerata  buds  grow  out  of  the  dorsal  surface.  The  operculum  is  discarded  at 


434  J.   F.   HARKKiAN   AND   1).   I.    AI.KOX 

TENTACLE 

BUD  VC-Y'-'     '5. 


EYE 


f 


CERAS 
BUD. 


40 


^FOOT 

FIGURE  2.  Newly  metamorphosed  Hermissenda.  A  pair  of  tentacle  huds  and  two  pairs  of 
cerata  buds  are  visible. 

metamorphosis.  The  newly-metamorphosed  animal  measures  about  400  /xm  in 
length  (Fig.  2).  The  body  is  still  divided  into  a  dorsal  visceral  mass  and  a 
ventral  foot,  and  the  larval  digestive  system  is  visible.  By  four  to  five  days 
post-metamorphosis  the  distinction  between  foot  and  visceral  mass  is  lost  and  the 
juvenile  has  begun  to  feed  on  hydroid  tissue. 

Metamorphosis  occurred  only  in  veligers  that  reached  full  development  be- 
tween days  34—58  post-hatching,  although  individuals  settling  after  day  50  soon 
died.  Although  larvae  may  survive  up  to  76  days,  there  was  little  or  no  shell 
growth  after  day  58. 

Diet 

Survival  of  Hermissenda  through  metamorphosis  was  low.  Addition  to  a  larval 
diet  of  Isochrysis  (jalhana  and  Monochrysis  littlicri  (5  ,11111  cell  diameter)  of  the 
larger  flagellate  Chroococcits  salina,  strain  3  C  (10-11  /xm  cell  length)  increased 
the  percentage  of  metamorphosis  from  1  to  5(/r .  Increasing  the  concentration  of 
Isochrysis  and  Monochrysis  did  not  improve  the  percentage  of  metamorphosis,  nor 
did  feeding  Chroococcns  alone. 

Post-metamorphie   stages,  however,   were  easily  maintained  in  the  laboratory. 


CULTIVATION  OF  HERMISSENDA 


435 


Variation  in  diet  significantly  affected  both  growth  rate  of  adults  and  number  of 
days  survived,  but  not  number  of  egg  masses  laid.  Diets  containing  tunicate, 
either  alone  or  in  combination  with  squid  and  mussel,  gave  the  best  growth  and 
survival  (Table  II). 

Initial  average  weights  of  four  groups  of  ten  small  wild  Hermissenda  each 
varied  from  299  nig  to  509  nig.  Average  weight  gains  on  each  of  the  four  diets 
were:  195.8  ±  362.1  mg  (squid)  ;  1218.0  ±  1514.3  mg  (mussel)  ;  2680.5  ±  1121.1 
nig  (tunicate);  and  2752.3  ±  1268.4  mg  (alternating  diet).  Animals  survived 
significantly  longer  on  tunicate-containing  diets  than  on  either  squid  or  mussel 
( t-test,  P  <  0.01,  df  ::  19).  Mean  number  of  days  survival  on  the  two  tunicate- 
containing  diets  was  63.9  days  (range  —  34—122  days). 

The  total  number  of  egg  masses  produced  did  not  vary  significantly  between 
diets  (Table  II).  There  was  no  significant  correlation  between  an  individuals 
growth  rate  on  any  diet  and  the  total  number  of  egg  masses  produced  by  that 
individual.  However,  there  was  a  significant  positive  regression  of  number  of  egg 
masses  produced  on  days  survived,  all  diets  combined  (polynomial  regression, 
P  <  0.01,  df  =  :  40)  (Fig.  3). 

Growth  rate  and  reproduction  in  fire  Pi   adults 

From  day  1  to  day  70  post-metamorphosis  increase  in  body  length  (on  a 
tunicate  diet)  was  approximately  linear,  averaging  0.82  ±0.11  mm  per  day.  The 
growth  rate  slowed  to  0.35  ±0.17  mm  per  day  between  days  71-120  post- 
metamorphosis.  The  largest  individual  attained  a  length  of  81.7  mm,  nearly 
equalling  the  length  of  the  largest  Hermissenda  obtained  from  the  field,  90  mm. 
After  clay  120  food  intake  decreased  and  the  animals  began  to  shrink.  Death 
occurred  between  116-137  days  post-metamorphosis  (X--  128  days). 

The  average  life-span  of  a  laboratory-reared  Hermissenda  encompasses  approxi- 
mately 163  days  (35  day  veliger  stage  plus  128  day  adult  stage),  confirming  that 
Hermissenda  is  a  subannual  species. 

Hermissenda  was  not  observed  to  self-fertilize.  Xo  egg  masses  were  de- 
posited by  Fl  adults,  which  were  maintained  separately,  until  three  animals  were 
allowed  to  copulate  on  day  65  post-metamorphosis  (total  egg  masses  --  28  from 
first  copulation  to  death).  Two  isolated  individuals  deposited  2-3  sterile  egg 
masses  each  between  days  95-133  post-metamorphosis. 

Fertile  egg  masses  were  deposited  in  the  laboratory  by  wild  specimens  of 
Hermissenda  as  small  as  73  mg,  and  motile  sperm  were  observed  in  squash  prepa- 

TABLE  II 
Growth  rate,  survival,  and  egg  mass  production  for  ten  specimens  of  Hermissenda  on  each  of  four  diets. 


Diet 

Average  growth  rate 
mg/day 

Average  days 
survived 

Average  egg  mass 
production 

Squid  mantle 
Mussel 
Tunicate 
Alternating 

10.1    ±  11.0 
31.4  ±  26.7 

58.4  ±  35.6 
55.3  ±  14.9 

28.0  ±  11.4 
45.4  ±  14.8 
65.0  ±  25.3 
62.9  ±     7.0 

2.0  ±   1.8 
3.7  ±  2.4 
3.5  ±  3.9 
1.9  ±  1.6 

436 


J.  F.  HARRK'iAX   AND  D.  L.  ALKON 


=  0.36l4-f  0.0493  X 


0.0 


28.0          56.0          84.0        112.0          140.0 
DAYS    SURVIVAL 

FIGURE  3.     Regression  of  number  of  egg  masses  deposited  on  days  survival. 

rations  from  wild  individuals  weighing  34  mg  (1.12  cm  body  length).  Egg 
production  in  individuals  from  wild  populations  is  estimated  to  begin  at  about 
1.5  months  post-metamorphosis  and  continue  until  death  at  5-8  g,  four  months 
post-metamorphosis.  Both  the  total  number  of  egg  masses  produced  and  the 
age  at  which  egg-laying  commences  depend  on  age  at  initial  copulation. 

DISCUSSION 

Hermissenda  crassicornis  is  one  of  several  nudibranch  species  which  have  been 
reared  through  metamorphosis  in  the  laboratory  (Bonar  and  Hadfield,  1974; 
Thompson,  1958,  1962,  1967;  Tardy,  1970;  Perron  and  Turner,  1977;  Harris, 
1975).  Harris  (1975)  and  Perron  and  Turner  (1977)  have  successfully  reared 
nudibranch  species  having  planktotrophic  (feeding)  larvae  from  egg  to  egg. 
Other  nudibranch  species  reared  have  been  either  lecithotrophic  or  direct  de- 
velopers. Hermissenda  has  a  longer  obligatory  planktotrophic  stage,  34  days,  than 
either  Phcstilla  melanobranchia  Bergh  1874  (Harris,  1975)  or  Doridclla  obscura 
Verrill  (Perron  and  Turner,  1977). 

The  length  of  the  veliger  stage  in  Hermissenda  is  similar  to  that  reported  for 
five  species  of  Pacific  aplysiid  opisthobranchs,  30-34  days  (Kriegstein,  Castellucci, 
and  Kandel,  1974;  Switzer-Dunlap  and  Hadfield,  1978).  Switzer-Dunlap  and 
Hadfield  (1978)  observed  a  plateau  in  shell  growth  before  metamorphosis  in  four 
aplysiid  species  similar  to  that  noted  in  Hermissenda.  As  adults,  three  of  the 
four  above  mentioned  aplysiid  species  were  reported  to  grow  to  a  larger  size 
in  the  laboratory  than  in  the  field.  No  specimens  of  Hermissenda  fed  in  the  laboratory 
have  exceeded  the  maximum  size  of  wild  individuals. 

Stages  in  the  life  cycle  of  Hermissenda  follow  in  the  same  sequence  as  the 
seven  general  life  history  stages  listed  by  Bonar  and  Hadfield  (1974)  :  hatching, 


CULTIVATION  OF  HERMISSENDA  437 

competency  to  metamorphose,  velum  loss,  shell  and  opercnlum  detachment  and 
loss,  and  sinking  of  the  visceral  mass  into  the  foot.  The  seventh  stage,  the 
pseudovermis  stage,  is  eliminated  in  metamorphosing  specimens  of  Hermissenda 
which  grow  tentacle  buds  and  cerata  buds  as  they  crawl  out  of  the  shell. 

The  life  cycle  of  Hcnuisscnda,  as  observed  in  the  laboratory,  follows  the  pat- 
tern described  for  other  hydroid-eating  nudibranchs  by  Thompson  (1964)  and 
Clark  (1975).  Animals  used  in  the  present  study  came  only  from  the  Monterey 
Bay  population ;  however,  reported  sizes  of  eggs  and  egg  capsules,  and  structure 
and  size  of  the  egg  masses  deposited  by  individuals  from  other  parts  of  the 
species  range  are  within  the  range  of  values  reported  here  (Hurst,  1967; 
O'Donoghue  and  O'Donoghus,  1922). 

The  most  variable  factor  observed  in  populations  of  veligers  and  adults  was 
growth  rate.  In  the  veliger  stage  part  of  this  variation  may  have  been  due 
to  culture  conditions.  Growth  of  larvae  may  have  been  inhibited  by  the  anti- 
biotic used,  chloramphenicol,  known  to  inhibit  protein  synthesis  in  eukaryotes  as 
well  as  bacteria  (Pestka,  1975),  or  the  larval  diet  may  have  been  suboptimal  for 
many  veligers. 

Large  laboratory  populations  of  Hermissenda  can  be  maintained  on  the  tunicate 
dona  intestinalis,  which  is  commonly  found  in  Cape  Cod  waters.  Year-round 
availability  of  dona  and  ease  of  collection  gives  it  an  advantage  over  the  normal 
field  diet,  which  consists  primarily  of  numerous  coelenterate  species,  as  well  as 
tunicates.  A  mixed  diet  did  not  markedly  improve  growth  or  survival  over  the 
single  item  tunicate  diet. 

In  Hermissenda,  individuals  which  have  the  fastest  growth  rates  are  also  the 
largest  adults.  A  program  of  selective  breeding  of  Hermissenda  will  concentrate, 
at  least  initially,  on  selection  for  fast  growth  rates.  High  selection  pressure  is 
already  exerted  on  the  laboratory  population  in  terms  of  survival  in  the  specific 
culture  conditions  utilized,  and  because  adults  are  reared  on  a  diet  of  only 
tunicate. 


We  would  like  to  thank  Richard  Waltz  for  assistance  with  the  statistical  pro- 
gramming, Helen  Stanley  of  Woods  Hole  Oceanographic  Institution  for  providing 
the  initial  algal  cultures,  Ruthanne  Theran  for  technical  assistance,  and  Dr.  Izja 
Lederhendler  for  his  critical  comments  on  the  manuscript. 

SUMMARY 

1.  Hermissenda  crassicornis  is  a  subannual  nudibranch  species  that  reproduces 
year-round. 

2.  There  is  a  significant  positive  relationship  between  adult  weight,  diameter 
of  the  egg  mass,  estimated  number  of  eggs  per  egg  mass,  and  average  number  of 
eggs  per  capsule. 

3.  There  is  a  planktonic  veliger  stage  of  34  days  minimum  at  13°-15°  C. 

4.  Larvae  metamorphose  on  at  least  three  species  of  hydroids. 

5.  To  develop  in  reasonable  numbers  to  a  state  competent  to  metamorphose 

; 


J.   I''.  HARRIGAN  AND   I).   L   AI.KON 

veligers  require  a  diet  that  includes  phytoplankton  of  larger  cell  size   (10-11 
than  the  commonly  used  Isochrysis  and  Monochrysis   (5  /AIH). 

6.  Although  Hermissenda  feeds  on  a  wide  variety  of  sessile  invertebrate  species 
in  the  ocean,  a  diet  of  tunicate  alone    (dona   intcstintilis)    promotes  good   growth 
and  survival  in  the  laboratory. 

7.  Egg  mass  deposition   is  initiated  only  after  first  copulation,  except  in   the 
last  month  of  life,  and  continues  from   about   one-month   post-metamorphosis   to 
death,  at  about   four  months  post-metamorphosis.     Generation  time    (egg-to-egg) 
may  lie  as  short  as  2.5  months. 

8.  A  laboratory  strain  of  Hermissenda  is  being  established  to  provide  animals 
of  known  history  for  research  on  the  neural  correlates  of  behavior.     Animals,  at 
least   initially,   are  being   selected   for   fast   growth   rate. 

LITERATURE  CITED 

ALKON,   D.    L.,    1974.      Associative    training   of    Hermissenda    crassicornis.      J.    Gen.    Physiol., 

64:  70-84. 
ALKON,    D.    L.,    1975.      Neural    correlates    of    associative    training    in    Hermissenda.      J.    Gen. 

Physio!.,  65  :   46-56. 
ALKON,   D.   L.,    1976.     Neural   modification   by   paired   sensory   stimuli.     /.    Gen.   Phvsio!.,   68: 

341-358. 
BIRKELAND,    C.,    1974.      Interactions    between    a    sea    pen    and    seven    of    its    predators.      Ecol. 

Momujr..  44:   211-232. 
BONAR,  D.  B.,  AND  M.  G.  HADFIELD,   1974.     Metamorphosis  of  the  marine  gastropod  Phcstilhi 

sibotjac     Bergh     (Nudibranchia :     Aeolidacea).       I.       Light    and    electron    microscope 

analysis  of  larval  and  metamorphic  stages.     J.  E.vp.  Alar.  Biol.  Ecol.,  16:  227-255. 
CLARK,   K.   B.,   1975.     Nudibranch   life  cycles   in  the   Northern  Atlantic  and  their   relationship 

to  the  ecology  of  fouling  communities   Hcli/o.    ll'iss.   Meeresunters,   27:   28-69. 
GUILLARD,    R.    L.,    1975.      Culture   of   phytoplankton   for   feeding   marine   invertebrates.      Pages 

29-71    in    W.    L.    Smith    and    M.    H.    Cbanley,    Eds.,    Culture    of    marine    iurcrtclinilc 

animals.   Plenum   Press,   New   York. 

HADERLIE,  E.  C.,  1968.     Marine  fouling  organisms  in  Monterey  Harbor.     Vcliycr.  10:  327-341. 
HADERLIE,   E.  C.,  J.  C.  MELLOR,  C.   S.  MINTER  III,  AND  G.  C.   BOOTH,   1975.     The   sublittoral 

bentbic   fauna   and    flora   off    Del    Monte    Beach,    Monterey,    California.      1'clitjcr,    17  : 

185-204. 
HARRIGAN,    J.    F.,    AND   D.    L.    ALKON,    1978.      Laboratory   cultivation   of    Haininoea    solitaria 

Say,    1822  and   Elysiti   chlorotica    Gould,    1870.      Vcliger,   in    press. 
HARRIS,    L.    G.,    1975.      Studies   on   the    life    history    of   two    coral-eating   nudibranchs    of   the 

genus  Phcstilhi.    Biol.  Bull.,  149:  539-550. 
HURST,   A.,    1967.      The   egg   masses   and   veligers   of   thirty    Northeast    Pacific   opisthobrancbs. 

Velificr.  9:   255-288. 
KRIEGSTEIN,   A.  R.,   V.   CASTELLUCCI,  AND  E.   R.    KANDEL,    1974.     Metamorphosis   of  Aplysia 

californicn    in    laboratory    culture.      Proc.    Nat.    Acad.    Sci.    USA,    71  :    3654-3658. 
LANCE,  J.  R.,  1966.     New  distributional  records  of  some  northeastern  Pacific  Opisthobranchiata 

(Mollusca:    Gastropoda)    with   descriptions   of   two   new   species.      Veligcr,   9:    69-81. 
MACFARLAND,  F.  M.,   1966.     Studies  of  Opisthobranchiate  Mollusks  of  the   Pacific   Coast   of 

North  America.     Calif.   Acad.   Sci.   Mem..   6:    1-546. 
O'DoNOGHUE,  C.  H.,  AND  E.  O'DONOGHUE,   1922.     Notes  on  the  nudibrancbiate  mollusca  from 

the  Vancouver  Island  Region.     II.     The  spawn  of  certain  species.      Trans.  Roy.   Can. 

hist.  XIV,   1  :    131-143. 

PERRON,   F.   E.,    AND    R.    D.    TURNER,    1977.      Development,    metamorphosis,    and    natural    his- 
tory   of    the    nudibrancb    Doridclla    ohscura    Verrill     (Corambidae:     Opisthobranchia) . 

/.  Exp.  Mar.  Biol.  Ecol.,  27  :  171-185. 
PESTKA,  S.,  1975.     Cblorampbenicol.     Pages  370-395  in  I.  W.  Corcoran  and  F.  E.  Habb,  Eds., 


CULTIVATION  OF  HERMISSENDA  439 

Mechanism  of  action  nf  antimicrobial  and  antitumor  ai/ents.  Springer-Verlag,  New 
York. 

SwiTZER-DuNLAP,  M.,  AND  M.  G.  HADFiELD,  1978.  Observations  on  development  and 
metamorphosis  of  four  species  of  Aplysidae  (Gastropoda,  Opisthobranchia)  in  lab- 
oratory culture.  J .  E.rp.  Mar.  Biol.  Ecol.,  in  press. 

TARDY,  J.,  1970.  Contribution  a  1'etude  des  metamorphoses  chez  les  nudibranches.  Ann.  Sci. 
Nat.  Zool.  Bio.  Anim..  Scr.  12.  T.,  12:  299-371. 

THOMPSON,  T.  E.,  1958.  The  natural  history,  embryology,  larval  biology  and  post-larval 
development  of  Adalaria  pro.ritini  (Alder  and  Hancock)  (Gastropoda,  Opistho- 
branchia). Phil.  Trans.  Roy.  Soc.  Loud.  Scr.  B..  242:  1-58. 

THOMPSON",  T.  E.,  1961.  The  importance  of  the  larval  shell  in  the  classification  of  the  Saco- 
glossa  and  Acoela  (Gastropoda,  Opisthobranchia).  Proc.  Malacol.  Soc.  Land.,  34: 
233-238. 

THOMPSON,  T.  E.,  1962.  Studies  on  the  ontogeny  of  Tritonia  liomberai  Cuvier  (Gastropoda, 
Opisthobranchia).  Phil.  Trans.  Roy.  Soc.  Loud.  Scr.  B..  245:  171-218. 

THOMPSON,  T.  E.,  1964.  Grazing  and  the  life  cycle  of  British  nudibranchs.  Pages  275-297  in 
D.  J.  Crisp,  Ed.,  Grazing  in  terrestrial  and  marine  environments.  Blackwell  Press, 
Oxford,  England. 

THOMPSON,  T.  E.,  1967.  Direct  development  in  a  nudibranch,  Cadlina  laeris,  with  a  dis- 
cussion of  developmental  processes  in  Opisthobranchia.  /.  Mar.  Biol.  Assoc.  U.K.. 
47  :  1-22. 

YARNALL,  J.  L.,  1972.  The  feeding  behavior  and  functional  anatomy  of  the  gut  in  the  eolid 
nudibranchs  Hennissenda  crassicornis  (Eschscholtz,  1831)  and  Aeolidia  papillosa 
(Linnaeus,  1761).  Ph.D.  dissertation.  Stanford  University,  134  pp.  (Diss.  Ahstr.. 
33B  (6)  :  2864,  order  number  72-30,725.) 


Reference:    />'/.»/.  />»//..  154:   440-452.  (June, 


CAPACITY    FOR    BIOSYNTHESIS    OF    PROSTAGLANDIN-RELATED 

COMPOUNDS:   DISTRIBUTION   AND   PROPERTIES   OF   THE 

RATE-LIMITING  ENZYME  IN  HYDROCORALS,  GOR- 

GONIANS,  AND  OTHER  COELENTERATES  OF 

THE  CARIBBEAN  AND  PACIFIC 

DANIEL  E.  MORSE,  MARK  KAYNE,  MARK  TIDYMAN,  AND  SHANE  ANDERSON 

Department  of  Biological  Sciences  and   The  Marine  Science  Institute, 
University   of   California,  Santa   Barbara,   California   93106 

The  search  fur  new  marine  sources  of  physiologically  potent  chemicals  of 
interest  to  biology,  and  potential  utility  to  medicine,  agriculture,  industry,  and  re- 
search, has  in  many  cases  been  hindered  by  the  lack  of  analytical  procedures  of 
sufficient  generality,  rapidity,  and  adaptability  to  use  in  the  field.  This  has  been 
particularly  so  in  the  case  of  the  hormone-like  prostaglandin-related  compounds 
(PGRCs),  which  are  now  known  to  include  a  large  and  confusing  multiplicity  of 
prostaglandins  (PCs),  prostacyclins,  thromboxanes,  and  prostaglandin-endo- 
peroxides  (Karim  and  Rao,  1975;  Hamberg,  Svensson  and  Samuelsson,  1975, 
Pace-Asciak  and  Wolfe,  1971  ;  Johnson,  Morton,  Kinner,  Gorman,  McGuire  and 
Sun,  1976). 

Very  high  levels  of  the  prostaglandins  PGE2,  PGA2,  and  certain  of  their 
related  isomers  have  been  found  in  different  clonal  populations  of  the  Caribbean 
gorgonian,  Plc.raiira  homomalla  (Anthozoa:  Gorgonacca)  (Weinheimer  and 
Spraggins,  1969;  Weinheimer,  1974;  Light  and  Samuelsson,  1972;  Schneider, 
Hamilton  and  Rhuland,  1972).  This  finding  generated  considerable  interest  in 
the  potentials  for  development  and  conservation  of  this  gorgonian  as  a  major 
medical  resource  (Bayer  and  Weinheimer,  1974),  although  commercial  interest 
in  this  fragile  and  slowly  growing  species  (Kinzie,  1974;  Hinman,  Anderson 
and  Simon,  1974;  Jordan,  Castanares  and  Ibarra,  1978)  has  been  supplanted  by 
recent  improvements  in  synthetic  methods  for  the  production  of  some  of  the  prosta- 
glandins. The  full  extent  of  the  distribution  and  potential  resource  of  PGRCs 
from  the  marine  invertebrates,  as  well  as  the  functions  of  the  PGRCs  in  these 
animals,  remain  largely  unknown,  however.  There  are  well  over  a  hundred 
prostaglandins  and  other  PGRCs  now  recognized,  with  newly  identified  members 
of  this  family  being  discovered  at  an  exponential  rate  (Karim  and  Rao,  1975). 
Assays  are  further  complicated  by  the  fact  that  these  compounds  are  for  the  most 
part  highly  unstable  (under  physiological,  aqueous,  and  aerobic  conditions),  and 
possess  overlapping  spectra  of  physical  and  biological  properties,  thus  necessitating 
resolution  and  analysis  by  complicated  and  specialized  techniques  (Schneider,  1976; 
Salmon  and  Karim,  1976). 

All  of  the  PGRCs,  however,  are  synthesized  from  a  common  (and  unstable) 
intermediate:  a  prostaglandin-endoperoxide  (PGEP)  (Hamberg  and  Samuelsson, 
1973,  1974).  The  enzyme-complex  catalyzing  the  rate-limiting  step  in  the  bio- 
synthesis of  this  central  intermediate  is  known  as  prostaglandin-endoperoxide  syn- 

440 


PROSTAGLANDIN  SYNTHETASE  IN  CORALS  441 

thetase  (also  known  as  prostaglandin  synthetase  or  fatty  acid  cyclo-oxygenase) 
(Miyamoto,  Ogino,  Yamanioto  and  Hayaishi,  1976).  Although  several  techniques 
for  the  assay  of  this  enzyme  are  available  (Samuelsson,  Granstrom,  Green,  Ham- 
berg  and  Hammarstrom,  1975;  Salmon  and  Karim,  1976),  use  of  these  assays  for 
the  direct  measurement  of  a  tissue's  maximal  capacity  for  PGEP  synthesis  (and 
thus,  the  total  capacity  for  subsequent  biosynthesis  of  PGRCs)  has  been  complicated 
by  the  pronounced  autocatalytic  and  autodestructive  activities  of  the  enzyme 
during  such  procedures  (Miyamoto  ct  al.,  1976;  Lands  and  Rome,  1976).  Such 
marked  deviations  from  simple  first-order  kinetics  result  in  a  complex,  nonlinear 
proportionality  of  the  reaction  with  respect  to  both  time  and  the  amount  of  enzyme 
present,  thus  limiting  the  usefulness  of  these  techniques  for  comparative  assess- 
ments of  relative  PGRC  biosynthetic  capacities. 

The  PGEP  synthetase-catalyzed  reaction  is  markedly  stimulated  by  hydrogen 
peroxide  (H2O2),  and  the  enzyme  from  many  invertebrate  sources  appears  to 
generate  this  activator  autocatalytically  during  the  course  of  its  normal  reaction 
(Morse,  Duncan,  Hooker  and  Morse,  1977,  1978).  Addition  of  exogenous  hy- 
drogen peroxide  (or  addition  of  a  hydrogen  peroxide  generating  system) 
rapidly  activates  PGEP  synthetase  ;  the  rate  of  the  reaction  catalyzed  by  this  enzyme 
is  then  easily  measured,  and  is  directly  proportional  to  the  amount  of  the  enzyme 
present.  Based  upon  this  finding,  a  rapid  and  convenient  spectrophotometric 
micro-assay  for  PGEP  synthetase  (Takeguchi  and  Sih.  1972)  was  modified, 
especially  adapting  it  for  use  in  the  field  by  inclusion  of  a  stable  enzymatic 
HoGVgenerating  system.  Using  this  technique  to  measure  the  levels  of  PGEP 
synthetase  in  a  variety  of  marine  coelenterates  from  the  Caribbean  and  Pacific, 
especially  high  specific  activities  of  this  enzyme  were  found  in  several  of  the 
plexaurid  Gorgonacca  (including  P.  honwuialla),  in  three  species  of  "Hydro- 
corallia"  (Millcporina  and  Stylastcrina),  and  in  two  species  of  Hydroida  ;  significant 
levels  of  the  enzyme  were  also  found  in  species  belonging  to  other  orders,  as  well. 

MATERIALS  AND  METHODS 

Specimen   collection 

Marine  coelenterates  were  obtained  from  both  the  Caribbean  ( Bonaire,  Nether- 
lands Antilles;  July-August,  1976;  1-30  m  depth)  and  the  Pacific  (Santa  Barbara 
Channel,  California;  August,  1976-July,  1977;  0-15  m  depth).  Small  samples 
of  tissue  (generally  <  5  g,  including  associated  skeletal  and  substrate  material) 
were  collected  from  these  marine  species  and  sealed,  in  situ,  in  sparate  polyethylene 
bags  of  sea  water  (ca.  100-200  ml')  ;  these  were  brought  ashore  for  prompt  assay. 
Freshwater  hydroids  were  obtained  from  the  Carolina  Biological  Corporation. 
Only  fresh,  live  specimens  were  used  for  all  assays  reported  here.  Identification 
and  classification  of  species  were  made  according  to  Hyman  (1940),  Bayer  (1961), 
Roos  (1971),  Boschma  (1956),  Smith  (1971),  Smith  and  Carlton  (1975),  Dur- 
ham and  Bernard  (1952),  Johnson  and  Snook  (1955),  and  Allen  (1976). 

Preparation  of  extracts 

All  samples  of  marine  species  were  rinsed  with  sea  water  after  separation 
from  associated  substrate  and  other  biota  as  necessary.  Tissue  was  removed  from 


442  MORSE,  KAYNK,  T1DVMAN  AND  ANDKRSON 

the  samples  ot  Scleractinia,  Milleporina  an<l  Stylasterina  by  scraping  the  skeletal 
material  with  a  scalpel,  and  irrigating  with  a  small  volume  of  chilled  /m-hydroxy- 
methylaminomethane  (Tris)-lICl  buffer  (10  HIM,  pH  7.1,  0°  C)  ;  all  other  samples 
were  minced  (at  0°  C)  to  facilitate  homogenization. 

Small  samples  of  weighed  tissue  (0.1-0.5  g)  were  homogenized  (0°  C)  in 
1-3  volume-equivalents  of  the  above  Tris-Cl  buffer,  using  a  small  glass  or  Tellon 
Bounce  homogenizer.  Participate  material  and  debris  were  removed  by  brief 
low-speed  centrifugation,  and  the  extracts  held  at  0°  C  for  immediate  assay. 

Assays 

The  catalytic  activity  of  PGEP  synthetase  was  measured  using  a  modification 
(Morse  ct  al.,  1977)  of  the  technique  originally  developed  by  Takeguchi  and  Sih 
(1972).  This  assay  spectrophotometrically  monitors  the  obligatory  co-oxidation 
of  the  colorless  aromatic  cofactor,  L-epinephrine,  as  it  is  converted  to  the 
intensely  red  adrenochrome  product.  The  assay-mixture  (1  ml,  20-23°  C)  con- 
tained Tris-Cl  buffer  (10  HIM,  pH  7.1  ),  arachidonic  acid  as  substrate  and  L-epine- 
phrine  as  cofactor  (each  at  1  HIM),  with  the  extract  to  be  assayed  and  other 
additions  as  indicated  in  the  text.  The  course  of  the  reaction  was  monitored  as  the 
rate  of  change  in  optical  absorbance  at  4SO  nm.  For  use  in  the  field,  assays  were 
performed  with  a  Bausch  and  Lomb  Mini-Spectrophotometer  (weight  ca.  200  g) 
and  stopwatch  ;  assays  performed  in  the  laboratory  made  use  of  a  Gilford  record- 
ing spectrophotometer.  Assays  of  the  same  homogenates  performed  in  parallel  with 
these  two  instruments  were  found  to  agree  within  ±  9'fi  . 

Where  indicated  (Table  I),  extracts  were  heated  at  90°  C  for  10  min  prior  to 
assay,  to  denature  enzyme  protein.  Also  as  indicated,  catalase  was  added  at 
0.1  /Ag/ml ;  phenylcyclopropylamine,  a:  pirin,  indomethacin,  acetaminophen,  DDTC, 
and  EDTA  were  added  at  1  HIM  concentration  as  shown. 

Aliquots  of  extracts  were  stored  frozen,  and  subsequently  assayed  for  protein 
concentration  by  the  method  of  Lowry,  Rosebrough,  Farr  and  Randall  (1951). 
Specific  activities  are  expressed  as  the  change  in  absorbance  (at  480  nm)  in  the 
assay  mixture  per  minute  per  mg  of  added  protein. 

Chemicals 

Tris  and  Tris-Cl  (pre-equilibrated  to  yield  pH  7.1  at  10  HIM,  22  '  C,  L-epine- 
phrine,  glucose  oxidase  and  catalase  were  obtained  from  the  Sigma  Chemical  Co. ; 
H2O2  (30^,  stablized)  was  obtained  from  Mallinckrodt,  and  diluted  just  before 
use.  Diethyldithiocarbamate  (DDTC,  sodium  salt)  and  ethylenediaminetetraacetic 
acid  (EDTA,  tetrasodium  salt)  were  from  Fisher  Chemical  Corporation;  all  other 
chemicals  were  reagent  grade.  All  solutions  were  prepared  with  distilled  water. 

RESULTS 

That  the  "prostaglandin  A2  synthetase  complex"  of  P.  homovnalla  is  activated 
by  1  M  NaCl  (Corey,  Washburn  and  Chen,  1973)  was  verified  in  this  study, 
using  the  spectrophotometric  assay  for  the  PGEP  synthetase  reaction  ;  this  activa- 
tion was  found  to  be  a  general  (although  somewhat  variable)  property  of  the 


PROSTAGLANDIN  SYNTHETASE  IX  CORALS 


TABLE  I 


443 


of  PGRP  synthetase  in  extracts  of  Allopora  porphyni.  PGEP  synthetase  activity  was 
assayed  in  5  pi  aliquots  of  a  freshly  prepared  extract  (8.8  mg  protein/ml)  of  Allopora  porphyra 
as  described  in  the  text,  with  alterations  as  specified.  Both  the  maximal  rate  of  the  enzyme-catalyzed 
reaction  (in  the  presence  of  1  M  NaCl)  and  the  initial  rate  (in  the  presence  of  0.6  HIM  H»O-t)  were 
measured;  results  are  the  averages  (±s.d.)  of  duplicate  determinations  normalized  to  values  obtained 
icith  the  respective  complete  assay  mixtures.  The  maxima!  rate  of  the  reaction  (-{-NaCl,  measured 
after  ca.  5  min)  corresponds  to  0.27  pinole  epinephrine  oxidized  per  minute;  the  initial  rate  in  the 
complete  system  with  H-iOz  was  0.24  pmole/min. 


Assay  mixture 

Relative 

activity  (%) 

Maximal  rate 
(with  NaCl) 

Initial  rate 
(with  H2O2) 

Complete  system 
Omit  activator  (NaCl  or  H,O,) 

1(10  db  8 
41  ±  5 

100 
2 

±  6 
±  1 

+  catalase 

0  ±  0 

0 

±  0 

+  2x  Extract 

198  ±  4 

208 

±  9 

Omit  extract 

0  ±  0 

0 

±  0 

+  Heated  extract 

0  ±  0 

0 

±  o 

Omit  arachidonic  acid 

38  ±  3 

54 

±  2 

+  Phenylcyclopropylamine 
-(-Aspirin 
+  Indomethacin 

3  ±  0 

84  ±  6 
44  ±  5 

0 
54 
16 

±  0 

=t  6 
±  2 

-(-Acetaminophen 
+  DDTC 

61  ±  2 
0  ±  0 

62 
4 

±  7 
±  1 

+EDTA 

133  ±  5 

140 

±  12 

enzyme  from  most  of  the  coelenterates  assayed.  No  such  salt-stimulation  of 
PGEP  synthetases  was  observed  in  active  extracts  from  marine  echinoderms, 
molluscs,  or  fishes,  however. 

As  the  spectrophotometric  assay  affords  a  means  for  continuously  monitoring 
the  progress  of  the  enzymatic  reaction,  the  effect  of  salt  upon  the  coelenterate 
PGEP  synthetase  coud  be  studied  in  more  detail.  Addition  of  NaCl  increases 
both  the  maximal  (autocatalytic)  rate  and  the  final  yield  of  the  reaction  by 
ca.  2—3  fold;  there  is  no  significant  effect  of  salt  on  the  slow  initial  rate,  however. 
Final  yield  of  the  reaction  is  limited,  in  part,  by  an  enzymatic,  autoinhibitory 
process,  and  not  by  depletion  of  substrate.  [Similarly  complex  autocatalytic  and 
autoinhibitory  processes  also  have  been  observed  in  kinetic  analyses  of  the  reaction 
catalyzed  by  PGEP  synthetases  from  a  variety  of  mammalian  sources  (e.g..  Lands 
and  Rome,  1976;  Miyamoto  ct  a/.,  1976).] 

The  data  in  Table  I  illustrate  the  properties  of  the  PGEP  synthetase  in  an 
extract  of  the  Pacific  "hydrocoral,"  Allopora  porph\ra  (Hydrozoa:  Stylasterina), 
a  species  especially  rich  in  this  enzyme.  Similar  properties  were  found  for  the 
PGEP  synthetases  in  extracts  of  P.  houioinalla  (Anthozoa  :  Gorgonacea),  Millepora 
spp.  (Hydrozoa:  Milleporina),  and  Sertnlaria  turgid  a  and  Hydractinia  unllcri 
(both  Hydrozoa:  Hydroida)  ;  thus,  the  data  in  Table  I  are  generally  representative 
of  the  PGEP  synthetases  from  those  coelenterates  which  contain  significant 
quantities  of  this  enzyme  (cf.  Table  III). 

As  seen  in  Table  I,  the  maximal  rate  of  the  autocatalytic  reaction  is  stimulated 
ca.  2.5-fold  by  1  M  NaCl.  By  continuously  recording  the  change  in  absorbance 


444  MORSE,  KAYNE,  TIDYMAN  AND  ANDERSON 

during  the  spectrophotometric  assay,  this  maximal  rate  of  the  salt-stimulated 
enzyme-catalyzed  reaction  can  be  determined  with  a  high  degree  of  accuracy,  and 
is  proportional  to  the  amount  of  extract  added.  This  same  maximal  rate  (±  15%) 
can  be  obtained — with  no  autocatalytic  lag — by  providing  hydrogen  peroxide  as 
activator  in  place  of  NaCl  in  the  complete  assay  mixture.  Maximal  stimulation  of 
the  enzyme  from  coelenterate  tissues  was  found  to  occur  at  approximately  0.6  mM 
H^Oo ;  this  is  close  to  the  value  of  0.3  HIM  previously  found  to  give  optimal 
stimulation  of  P(iEP  synthetase  from  eggs  of  the  abalone,  Haliotis  rujcsccns 
(Morse  ct  al.,  1977,  1978).  As  expected,  the  peroxide-stimulated  reaction  is  com- 
pletely inhibited  by  the  addition  of  purified  catalase,  an  enzyme  which  rapidly  and 
specifically  decomposes  the  added  H^Oo  to  water  and  oxygen.  More  significant, 
however,  is  the  observation  that  both  the  autocatalytic  activation,  and  all  catalytic 
activity,  seen  in  the  absence  of  exogenous  peroxide  (±  NaCl)  are  completely 
inhibited  by  a  small  concentration  of  catalase.  This  observation,  also  made  with 
the  PGEP  synthetase  from  other  marine  invertebrates  (Morse  et  al.,  1977,  1978), 
indicates  that  both  the  activity  and  autocatalytic  activation  of  the  enzyme  from 
these  sources  normally  depend  upon  the  (autocatalytic)  generation  of  H^Oo  by  the 
PGEP  synthetase  itself. 

Both  the  maximal  rate  of  the  salt-stimulated  reaction  and  the  initial  (—  maxi- 
mal) rate  of  the  peroxide-stimulated  reaction  are  absolutely  dependent  upon  a  heat- 
labile  factor  (presumably  enzyme)  in  the  added  extract  (Table  I).  Dependence 
upon  the  added  substrate,  arachidonic  acid,  is  only  partial  and  widely  variable  from 
extract  to  extract,  presumably  reflecting  the  variable  presence  of  endogenous  lipid 
substrates  in  the  crude  extracts.  Enzymatic  activity  in  the  presence  of  either 
NaCl  or  H;>O:>  is  inhibited  to  various  extents  by  the  pharmacological  anti-inflam- 
matory, analgesic  and/or  antipyretic  agents  phenylcyclopropylamine,  aspirin, 
indomethacin,  and  acetaminophen  ;  these  agents  are  known  to  inhibit  PGEP  synthe- 
tases  from  a  variety  of  different  organisms  and  tissues  with  widely  varying  efficien- 
cies (Lands  and  Rome,  1976).  Phenylcyclopropylamine  is  most  efficient,  of 
these,  at  inhibiting  the  coelenterate  enzyme;  it  had  been  observed  previously  that 
the  salt-activated  enzyme  from  P.  Jwinoinalla  was  relatively  insensitive  to  indo- 
methacin, although  lower  concentrations  of  that  agent  than  used  in  the  present  study 
had  been  employed  (Corey  ct  al.,  1973).  Our  data  indicate,  however,  that  the 
initial  rate  of  the  reaction  catalyzed  by  the  peroxide-activated  coelenterate  enzyme 
is  significantly  more  sensitive  to  inhibition  by  aspirin  and  indomethacin  than  is  the 
maximal  rate  achieved  after  autocatalytic  activation  in  the  presence  of  salt.  As 
with  the  PGEP  synthetases  from  other  sources  (Morse  et  al.,  1977;  Letellier, 
Smith  and  Lands,  1973),  the  coelenterate  enzyme  is  strongly  inhibited  by  diethyl- 
dithiocarbamate  (DDTC),  a  chelator  strongly  specific  for  copper.  The  addition 
of  EDTA,  a  chelator  which  is  specific  for  heavy  metals  other  than  copper,  results 
in  a  slight  but  significant  increase  in  catalytic  activity.  These  latter  observations 
suggest  that  copper  may  play  some  essential  role  in  the  coelenterate  PGEP  synthe- 
tase, as  it  does  in  many  other  oxygenases  (Morse  ct  al..  1978)  ;  traces  of  other 
heavy  metals  appear  to  cause  slight  inhibition,  which  may  be  relieved  by  addition 
of  EDTA. 

The  peroxide-stimulated  reaction  was  further  adapted  for  use  in  an  assay 
which  could  be  performed  conveniently  under  field  conditions,  by  replacement  of 


PROSTAGLANDIN  SYNTHETASE  IN  CORALS  445 

the  H2O2  with  a  stable,  enzymatic  H2O2-generating  system.  As  seen  in  Table  II, 
a  simple  enzymatic  system  (consisting  of  glucose  oxidase  and  its  substrate,  D- 
glucose)  can  be  incorporated  in  the  spectrophotometric  assay  for  the  continuous 
production  of  H2O2  (and  glucuronic  acid)  in  situ.  The  purified  and  concentrated 
glucose  oxidase,  which  is  inexpensively  available  from  several  commercial  sources, 
proves  to  be  fairly  stable ;  such  preparations  can  tolerate  several  weeks  in  transit 
without  refrigeration,  with  little  significant  loss  in  activity.  Using  the  "coupled 
assay"  shown  in  Table  II,  with  optimal  concentrations  of  glucose  and  glucose 
oxidase  replacing  the  direct  addition  of  H2O2,  the  measured  activity  was  found 
to  exhibit  dependence  upon  added  coelenterate  extract  and  substrate,  and  sensi- 
tivity to  inhibitors,  closely  parallel  to  results  obtained  with  the  simple  H2O2-stim- 
ulated  reaction  shown  in  Table  I. 

Using  this  convenient  and  readily  portable  assay  procedure,  the  relative  PGEP 
synthetase  levels  were  measured  in  extracts  of  fresh,  live  tissue  from  a  variety 
of  coelentrates  of  the  Caribbean  and  eastern  Pacific  (Table  III).  In  addition  to 
the  very  high  levels  of  this  enzyme  found  in  P.  hoinoinalla,  high  or  significant  specific 
activities  were  found  in  four  other  species  of  plexaurids  and  Gorgonia  vcntaVma 
(all  Gorgonacea),  the  solitary  Scleractinian,  Cocnocyathus  bou'crsi,  the  Caribbean 
Antipatharian  ("black  coral")  Antipathes  atlantica,  the  Hydroids  Hydractinia 
inilleri  and  Sertiilaria  tnrgida.  and  four  species  of  Milleporina  and  Stylasterina  (the 
"Hydrocorallia").  Enzymatic  activities  from  all  of  these  sources  were  found  to  be 
dependent  upon  H2O2.  The  low  specific  activities  measured  in  the  other  species 
assayed  actually  reflect  lower  concentrations  of  the  enzyme,  rather  than  the 
presence  of  some  inhibitor  of  its  activity,  as  no  significant  inhibition  was  detected 
upon  mixing  any  of  the  extracts  (of  all  species  tested)  with  extracts  of  Plcxaura, 
Millcpora,  or  Allopora. 

The  distribution  of  the  enzyme  in  the  plexaurid  Gorgonacea  and  in  the  "Hydro- 
corallia" (Milleporina  and  Stylasterina)  appears  to  be  of  some  general  significance, 
although  few  other  taxonomic  or  physiological  correlates  of  this  distribution  are 
apparent.  It  should  be  noted  that  if  Muricca  is  included  in  the  Plexauridae, 
as  according  to  Bayer  (1961),  high  levels  of  PGEP  synthetase  may  not  be 
entirely  characteristic  of  this  family ;  however,  others  have  classified  this  genus 

TABLE  II 

"Coupled  assay"  with  endogenous  generation  of  HzOz  by  glucose  oxidase.  PGEP  synthetase  activity 
was  assayed  in  an  extract  of  Allopora  porphyra  as  described  in  Table  I,  except  that  the  otherwise 
complete  system  contained  "activator"  as  specified.  Results  are  the  averages  (±s.d.)  of  duplicate 
determinations,  normalized  to  the  value  obtained  in  the  presence  of  H^Oz  at  optimal  concentration. 

Initial  Rate 
Activator  (%) 

HiO2  (0.6  nut)  100  ±  1 

None  ±  1 

Glucose  (5  mM)  2  ±  1 

Glucose  Oxidase  (10  /ig)  14  ±  6 

Glue.     (5  mM)  +  Glue.  Ox.  (10  ^g)                                     115  ±  3 

Glue.  (50  mM)  +  Glue.  Ox.  (10  /xg)                                      82  ±  7 

Glue.     (5  mM)  +  Glue.  Ox.  (100  Mg)                                        83  ±  6 


446 


MOKSK,   KAYNE,  TIDVMAN   AM)  ANDERSON 


TAHLK   III 


Species-distribution  of  PtlEP  syuthetase  in  coelenterates.  Specimens  were  collected  and  assayed  as 
indicated;  specific  activities  are  the  averages  (±s.d.)  of  results  from  two  or  more  separate  colonies 
measured  in  the  "coupled"  assay  with  endogenous  generation  of  H-iO-i  (5  HIM  glucose  -\-  10  ng/ml 
glucose;  cf.  Table  II).  Collection  sites  are:  P,  Pacific;  C,  Caribbean;  f.w.,  freshwater.  An  asterisk 
denotes  species  with  high  specific  activity  of  PGEP  synthetase.  In  all  cases  in  which  significant  activity 
(>(>./)  was  detected,  proportionality  of  activity  with  added  extract,  and  dependence  upon  H-iO-i  were 
determined  (as  in  Tables  I  and  II).  Absence  of  inhibitors  in  all  extracts  was  verified  as  described 
in  the  text. 


Family 

Species 

Collection 
Site 

PGEP  Synthetase 
(Specific  Activity) 

(Anthozoa  :  Octocorallia) 

Plexauridae 

Plexaura  homomalla 

(var.  homomalla) 

C 

*     9.0  ±   1.8 

(var.  kiikenthali) 

C 

*    9.0  ±  2.2 

Plexauridae 

Plexaura  flexuosa 

C 

*    8.5  ±  1.6 

Plexauridae 

Pseudoplexaura  flagellosa 

C 

*    5.8  ±  1.5 

Plexauridae 

Eunicea  tourneforti 

(var.  tourneforti) 

C 

*    3.0  ±  1.2 

(var.  atra) 

C 

*    3.2  ±  0.3 

Plexauridae 

Plexaurella  dichotema 

C 

*    2.0  ±  0.6 

Gorgoniidae 

Gorgonia  ventalina 

C 

*     1.6  ±  0.3 

Gorgoniidae 

Pseudopterogorgia  america  na 

C 

<().!  ±  0 

Gorgoniidae 

Eugorgia  rubcns 

P 

0.3  ±  0.1 

Gorgoniidae 

Lophogorgia  chilensis 

P 

<().!  ±  0 

Gorgoniidae 

Fili  gel-la  mitsukurii 

P 

<().!  ±  0 

Muriceidae 

Muricea  calif  arnica 

P 

<().!  ±  0 

Clavulariidae 

Clavularia  sp. 

P 

0.3  ±  0.1 

Virgulariidae 

Stylatula  elongata 

P 

<().!  ±  0 

Virgulariidae 

A  ca  nthoptilurn  gracile 

P 

0.2  ±  0.1 

Renillidae 

Renilla  kollikeri 

P 

<0.1  ±  0 

(Anthozoa:  Hexacorallia) 

Seriatoporidae 

Madracis  decactis 

C 

0.2  ±  0.1 

as  belonging  to  a  separate  group  (as  indicated  in  Table  III).  No  significant 
differences  were  observed  between  two  subspecies  each  of  P.  homomalla  (var. 
homomalla  and  var.  kiikentliali;  Table  III),  Eunicea  tourneforti  (var.  tourneforti 
and  var.  atra;  Table  III),  or  Allopora  porphyra  (vars.  red  z<s.  orange;  cf. 
Ostarello,  1973),  when  these  pairs  were  collected  and  assayed  in  parallel.  Similarly, 
no  significant  differences  were  observed  (in  parallel  collections  and  assays)  be- 
tween PGEP  synthetase  levels  in  male  and  female  colonies  of  dioecious  species 
such  as  Plexaura,  Millepora,  or  slllopora. 

Corey  and  Washburn  (1974)  had  previously  shown  that  the  PG  synthetase 
complex  of  P.  houioinalla  resides  in  the  tissue  of  the  animal,  rather  than  in  its 
symbiotic  zooxanthellae.  In  view  of  the  suggestions  made  by  them  and  others 
(Corey  and  Washburn,  1974;  Gonzalez,  1978)  that  photosynthetic  products  of  the 
zooxanthellae  may  neverthless  contribute  to  (or  control)  the  biosynthesis  of 
PGRCs  in  coelenterates,  it  was  of  interest  to  determine  the  relative  activities  of 
PGEP  synthetase  from  colonies  of  the  same  species  exposed  to  widely  differing 
regimes  of  illumination.  However,  we  have  found  that  colonies  of  Millepora 
alcicornis  collected  from  depths  of  1  m  and  30  m  (the  extremes  of  its  depth- 


PROSTAGLANDIN  SVXTHETASE  IN  CORALS 


447 


distribution  which  we  observed),  when  assayed  in  parallel,  showed  no  significant 
differences  in  specific  activity,  thus  suggesting  that  photosynthetic  activity  may 
have  little  direct  influence  over  the  synthesis  or  activity  of  the  rate-limiting 
enzyme,  PGEP  synthetase. 

DISCUSSION 

In  their  studies  of  the  PG  synthetase  from  P.  hoinoiiialla,  Corey  ct  nl.  (1973) 
found  apparently  complete  dependence  of  activity  upon  added  NaCl,  whereas  our 
assays  detect  only  a  2-3  fold  stimulation  in  extracts  of  this  and  other  coelenterates. 
Possible  reasons  for  the  difference  between  these  observations  include  the  fact  that 
Corey  ct  al.  measured  the  final  yield  of  the  overall  enzymatic  synthesis  of  PGA2, 
whereas  we  have  measured  the  rate  of  the  reaction  catalyzed  by  PGEP  synthetase 
alone.  Also,  Corey  ct  al.  measured  the  final  cumulative  activity  in  extracts  which 
had  been  stored  frozen,  whereas  our  assays  were  performed  with  specimens  which 
had  been  freshly  collected  and  live  immediatly  prior  to  assay.  In  fact,  the  activity 
of  the  PGEP  synthetase  complex  was  found  in  this  study  to  be  only  partially 
stable  in  frozen  tissues,  with  samples  variably  losing  50-80%  of  their  activity 
when  kept  at  -  30°  C  for  two  months. 

TABLE    III — Continued 


Family 

Species 

Collection 
Site 

PGEP  Synthetase 
(Specific  Activity) 

Acroporidae 

Acropora  palmata 

C 

<().!  ±  0 

Agariciidae 

Agaricia  agaricites 

C 

0.2  ±  0 

Agariciidae 

Agaricia  fragilis 

C 

0.2  ±  0.1 

Faviidae 

Diploria  labyrinth  if  arm  is 

C 

<().!  ±  0 

Trochosmiliidae 

Meandrina  meandrites 

C 

<().!  ±  0 

Trochosmiliidae 

Dcndrogyra  cylindrus 

C 

0.3  ±  0.1 

Eupsammidae 

Ballanophyllia  elegans 

p 

0.3  ±  0.1 

Astrangidae 

Astrangia  lajollaensis 

p 

0.5  ±  0.2 

Caryophylliidae 

Coenocyathus  bower  si 

p 

*    3.3  ±  0.4 

Anthopleuridae 

A  iithopleura  elegantissinia 

p 

<().!  ±  0 

Anthopleuridae 

A  ntho  pleura  xa  nthogra  m  m  ica 

p 

<().!  ±  0 

Actiniidae 

Tealia  crassicornis 

p 

<().!  ±  0 

Sagartidae 

Corynactis  calif  or  n  ica 

p 

<().!  ±  0 

Antipathidae 

A  ntipathes  atlantica 

C 

*    3.7  ±  0.5 

Antipathidae 

A  ntipathes  rhipidion 

C 

0.2  ±  0.1 

(Hydrozoa) 

Bougainvilleidae 

Hydractinia  milleri 

p 

*    9.8  ±  2.1 

Tubulariidae 

Tubularia  crocea 

p 

<().!  ±  0 

Eudendriidae 

Eiidendrium  californicum 

p 

0.2  ±  0 

Hydridae 

Pelmatohydra  pseudoelegactis 

f.w. 

<().!  ±  0 

Hydridae 

Chlorohydra  I'iridissima 

f.w. 

<().!  ±  0 

Sertulariidae 

Sertularia  turgida 

P 

*  11.2  ±  2.9 

Campanulariidae 

Clytia  bakeri 

P 

<().!  ±  0 

Plumulariidae 

Aglaophenia  struthionides 

P 

<().!  ±  0 

Milleporidae 

Millepora  alcicornis 

C 

*  10.3  ±  2.4 

Milleporidae 

Millepora  complanata 

C 

*    8.8  ±  3.0 

Milleporidae 

Millepora  sqiiarrosa 

C 

*    8.3  ±  4.3 

Stylasteridae 

Allopura  porphyra 

1' 

*    4.6  ±  2.2 

Chondrophorae 

Velella  velella 

p 

<().!  ±  0 

448  MOKSK,   KAYNE,  TIDYMAN  AM)  ANDERSON 

The  PGEP  synthetase  reaction  stimulated  by  salt  remains  autocatalytic,  and 
thus,  difficult  to  measure;  assays  monitoring  the  yield  of  PG  products  have  proven 
unreliable  for  accurate  and  comparative  quantitations  of  enzymatic  activity  (Corey 
ft  al.,  1973;  Samuelsson  cf  a!.,  1975;  Miyamoto  ct  al.,  1976;  Lands  and  Rome, 
1976).  Using  the  spectrophotometric  assay  with  a  continuously  recording  spectro- 
photometer,  however,  reliable  determinations  of  the  rate  of  the  autocatalytic,  salt- 
stimulated  reaction  catalyzed  by  PGEP  synthetase  in  coelenterate  extracts  were 
obtained.  This  maximal  rate  is  directly  proportional  to  the  amount  of  extract 
added  (Table  I),  and  is  thus  useful  for  comparative  quantiations  of  enzyme 
activity. 

Previous  work  from  this  laboratory  has  demonstrated  that  the  PGEP  synthetases 
from  a  variety  of  marine  invertebrates  can  be  activated  by  hydrogen  peroxide ; 
this  activation  proceeds  with  immediate  elimination  of  the  autocatalytic  lag  in  the 
PGEP  synthetase-catalyzed  reaction,  and  thus  makes  possible  the  convenient 
quantitation  of  the  enzyme  with  simple  first-order  kinetics  (Morse  ct  al.,  1977, 
1978).  Similar  activation  (with  HoO^  in  place  of  NaCl ;  see  Table  I)  makes 
possible  the  direct  and  convenient  quantitation  of  the  enzyme  from  a  wide  variety 
( if  marine  coelenterates. 

That  H2O2  is  apparently  generated  by  tbe  enzyme  reaction  itself,  and  is  thus 
responsible  for  the  autocatalytic  activation  (observed  in  the  absence  of  added 
peroxide),  is  indicated  by  the  finding  that  the  addition  of  catalasc  (0.1  jug/ml)  to 
the  reaction-mixture  (±  1  M  NaCl)  completely  eliminates  both  autocatalytic 
activation  and  all  catalytic  activity  of  the  enzyme  in  extracts  of  the  coelenterates 
Plc.vatira,  Pseudoplexaura,  Antipathes,  Millcpora,  and  Allopora.  Similar  evi- 
dence has  been  found  for  the  enzyme  from  marine  molluscs  and  echinoderms 
(Morse  ct  al.,  1977,  and  unpublished  observations),  and  thus  appears  to  reflect 
a  general  property  of  the  reaction-mechanism  of  this  enzyme  from  many  inverte- 
brate species.  A  role  for  copper  at  the  active  site  of  these  enzymes  has  been 
postulated  in  the  generation  of  H2Oo  (Morse  ct  al.,  1978),  and  is,  in  part,  supported 
by  the  sensitivity  of  these  enzymes  to  the  copper-chelator,  DDTC  (Table  I ; 
Morse  ct  al.,  1977,  1978).  In  these  respects,  as  well  as  in  the  relatively  low  sensi- 
tivities to  the  anti-inflammatory  drugs  which  are  potent  inhibitors  of  the  mammalian 
enzymes,  the  properties  of  the  PGEP  synthetases  from  the  marine  invertebrates 
differ  from  those  of  the  enzymes  from  mammalian  sources  (see  also  Corey  ct  al., 
1973). 

From  a  practical  point  of  view,  there  are  several  advantages  which  use  of  the 
peroxide-stimulated  reaction  affords  over  measurement  of  the  salt-stimulated 
reaction.  Accurate  measurements  of  the  maximal  autocatalytic  rate  of  the  salt- 
stimulated  reaction  require  sophisticated  electronic  equipment  for  continuous  moni- 
toring and  recording  of  the  spectrophotometric  assay.  In  contrast,  the  initial  rate 
of  the  (first-order)  peroxide-stimulated  reaction  can  be  measured  readily  in  the 
field,  with  a  simple  spectrophotometer  (or  colorimeter)  and  stopwatch.  HoOo 
itself  is  unstable  in  dilute  solution,  and  in  concentrated  form  (or  as  the  solid, 
e.g.,  sodium  peroxide)  is  both  caustic  and  potentially  explosive,  and  thus  subject 
to  internationally  regulated  precautions  in  transport.  However,  the  peroxide- 
stimulated  reaction  can  be  further  adapted  to  use  in  the  field  by  replacement  of 
H2OL»  with  a  stable  enzymatic  H2O2-generating  system  (Table  II).  When  used 


PROSTAGLANDIN  SYNTHETASE  IN  CORALS  449 

with  miniaturized  and  highly  portable  spectrophotometric  equipment,  this  pro- 
cedure makes  convenient  and  reliable  assays  under  field  conditions  possible,  allowing 
comparisons  of  the  specific  activities  of  PGEP  synthetase  from  live,  freshly  collected 
specimens  of  a  variety  of  coelenterates  from  the  Caribbean  and  eastern  Pacific. 

Use  of  these  procedures  has  confirmed  the  identification  of  Plc.vaiini  Iioiiioinalla 
as  a  species  exceptionally  rich  in  PGEP  synthetase  (Table  III ;  Corey  el  al., 
1973;  Weinheimer  and  Spraggins,  1969;  Bayer  and  Weinheimer,  1974).  In 
addition,  this  study  has  identified  several  related  plexaurids,  as  well  as  certain 
other  Gorgonacea,  "Hydrocorallia",  Antipatharia,  Scleractinia,  and  Hydroida  as 
species  warranting  further  investigation  as  sources  of  potentially  great  PGRC 
biosynthetic  activity.  Although  little  systematic  pattern  is  discernible  in  the 
distribution  of  the  high  levels  of  PGEP  synthetase  observed,  it  may  be  significant 
that  all  of  the  hydrocoral  species  tested  (three  Millcpora,  one.  St\lastcrina  )  were 
found  to  have  exceptionally  high  levels  of  this  enzyme.  Since  the  total  productivity 
of  Pacific  and  Atlantic  species  (particularly  of  the  tropical  hydrocorals)  thus  identi- 
fied far  exceeds  the  relatively  low  productivity  of  the  Caribbean  gorgonian  P. 
hoinoiiialla  (Hinman,  1974;  Jordan,  et  al.,  1978),  these  findings  may  serve  to 
relieve  and  diversify  pressure  for  exploitation  upon  this  latter  and  potentially 
threatened  species. 

Marine  coelenterates  are  the  phylogenetically  simplest  organisms  in  which 
significant  levels  of  PGEP  synthetase  thus  far  have  been  found.  Such  activity 
was  not  detected  in  several  species  each  of  freshwater  Protozoa  and  marine  Porifera. 
Specific  activities  of  enzyme  in  the  most  active  coelenterate  extracts  (Table  III) 
exceed  those  found  in  mammalian  reproductive  tissues,  although  they  are  about 
50%  lower  than  the  highest  values  found  in  the  eggs  of  abalone,  Haliotis  spp., 
and  the  urchins,  Strongyloccntrotns  and  Lytechinus  spp.  (Morse  et  al.,  1977, 
and  unpublished  observations).  Although  data  implicate  this  enzyme  in  the  control 
of  reproductive  processes  in  both  abalones  (Morse  et  al.,  1977,  1978)  and  urchins 
(Jensen  and  Morse,  unpublished  observations),  there  is  as  yet  no  information 
regarding  the  physiological  functions  of  the  especially  active  PGEP  synthetases 
of  the  marine  coelenterates.  Similarly,  the  final  (PGRC)  products  of  the  enzyme 
from  these  sources,  with  the  exception  of  those  from  P.  hoiiioinalla,  remain  to  be 
identified. 

The  apparent  distribution  of  PGEP  synthetase  activity  found  in  the  marine 
coelenterates  (Table  III)  might  reflect  some  pattern  of  seasonal  variation,  per- 
haps in  reproductive  or  other  -specialized  functions  and/or  tissues.  However,  no 
such  seasonal  variation  has  been  detected  in  samples  of  five  of  the  Pacific  species 
(Allofiora,  Sertuhiria,  Lophogorgia,  Muricca,  and  Tealui)  collected  and  assayed 
at  intervals  throughout  the  year.  It  is  possible,  then,  that  the  high  levels  of 
PGEP  synthetase  characteristic  of  certain  species  may  reflect  a  role  in  some 
fundamental  process  such  as  the  regulation  of  ion-  and  water-transport,  as 
originally  suggested  by  Christ  and  Van  Dorp  (1972). 

Alternatively,  the  potent  PGRCs  in  these  species  might  play  some  role  in 
defense  against  predation  or  parasitism,  or  in  specialized  aggressive  or  prey- 
securing  functions.  The  effectiveness  of  the  PGRCs  from  molested  and  damaged 
colonies  of  P.  huniuinalla  in  causing  severe  irritation  and  other  symptoms  of 
intoxication  in  human  collectors  has  been  documented  previously  (Brooks  and 


450  MORSE,  KAYNK,  TIDYMAN  AND  ANDKRSON 

White.  1974).  For  a  discussion  of  the  many  physiological  functions  in  which 
postaglandins  and  PGRCs  have  heen  implicated,  the  reader  is  referred  to  tin- 
recent  comprehensive  reviews  edited  by  Karim  (1975.  1976). 


This  research  was  supported,  in  part,  hy  the  Marine  Science  Institute  of  the 
University  of  California  at  Santa  Barbara.  We  gratefully  acknowledge  the  excel- 
lent technical  assistance  of  Aileen  Morse,  Helen  Duncan,  and  the  divers  of  the 
Department  of  Biological  Sciences,  as  well  as  the  generous  assistance  of  Dr.  Henry 
Ofren,  Fran  Ciluaga,  and  the  entire  staff  of  the  Marine  Science  Institute. 

We  also  wish  to  thank  Drs.  Ingvaar  Kristiansen,  Hans  De  Kruijf,  and  Rolf 
Bak  of  the  Caraihisch  Marien  Biologisch  Instituut,  Curacao;  Mr.  L.  D.  Gerharts, 
Director  of  the  Kraalendjik  Trading  Company;  and  Captain  Don  Stewart  and  his 
divers  Ebo,  Tony,  Eddie  and  Adi,  of  Aquaventure,  Bonaire,  for  their  warm 
hospitality  and  most  generous  assistance. 

Portions  of  this  research  were  conducted  at  the  CARMAHI  field  station  at  Malmok, 
and  at  Captain  Don  Stewart's  Aquahabitat,  Bonaire,  for  which  facilities  we  are 
very  grateful. 

SUMMARY 

A  convenient  and  reliable  assay  is  described  for  PGEP  synthetase,  the  rate- 
limiting  enzyme  determining  the  total  capacity  for  biosynthesis  of  prostaglandin- 
related  compounds.  Results  of  such  assays,  performed  with  fresh  specimens  under 
both  field  and  laboratory  conditions,  newly  identify  several  marine  coelenterate 
species  as  potentially  important  resources  of  PGRCs  for  research  and  possible  de- 
velopment. Properties  of  the  typical  marine  coelenterate  PGEP  synthetase,  and 
the  reaction  which  this  enzyme  catalyzes,  have  been  further  characterized. 

LITERATURE  CITED 

ALLEN,  R.  K.,  1976.     Common  intertidal  invertebrates  of  southern  California.     Peek,  Palo  Alto, 

316  pp. 
BAYKR,   F.   M.,    1961.      The  shallou'-n'atcr   Octocorallia   of   the    U'est   Indian    rei/ion.      Martinus 

Nijhoff,  The  Hague,  373  pp. 
BAYER,    F.    M.,    AND   A.   J.    WEINHEIMER,    1974.      Prostaglandins   from    Plexaura   homomalla : 

ccoh>(/y,  utilization  and  conservation   of  a  major  medical  marine  resource.     University 

of  Miami   Press,  Coral   Gables,   Florida,   165  pp. 
BOSCIIMA,    H.,    1956.      Milleporina    and    Stylasterina.      Pages    90-106    in    R.    C.    Moore,    Ed., 

Treatise    invertebrate    paleontology.      Geological    Society    of    America,    University    of 

Kansas   Press,   Manhattan. 
BROOKS,  C.  D.,  AND  G.  J.  WHITE,  1974.     Intolerance  patterns  in  Plerura  homomalla  collectors: 

case  reports  and  diagnostic  studies.     Pages  127-136  in  F.   M.  Bayer,  and  A.  J.   Wein- 

heimer,   Eds.,  Prostaglandins  from    Plexura  homomalla :   ecology,   utilization   and   con- 
servation   of   a    major   medical    marine    resource.      University    of    Miami    Press,    Coral 

Gables,    Florida. 
CHRIST,  E.  J.,  AND  D.  A.  VAN  DORP,  1972.     Comparative  aspects  of  prostaglandin  biosynthesis 

in  animal  tissues.     Biochim.  Kiophys.  Acta,  270  :  537-545. 
COREY,   E.   J.,    AND   W.    N.    WASHBURN,    1974.      The    role   of   the   symbiotic   algae   of   Plc.raura 

lioinoni  ilia   in  prostaglandin  biosynthesis.     J.  Am.   Chem.   Soc.,  96:   934-935. 
COREY,   E.   J.,   W.   N.  WASHUVRN,   AND  J.   C.   CHEN,   1973.      Studies  on   the   prostaglandin   AL. 

synthetase   complex   from   Plc.raura   homomalla.    J.    Am.    Chem.   Soc.,   95:    2054-2055. 


1'ROSTAGLANDIX  SYXTHKTASK   IX   CORALS  451 

DURHAM,  J.  W.,  AMI  J.  L.  BERNARD,  1952.  Stony  corals  of  the  (.-astern  Pacific  collected  by  the 
Valero  III  and  Valero  IV.  Allan  Hancock  Pacific  Expedition.  16:  1-47. 

GONZALEZ,  P.  A.  B.,  1978.  The  hehavior  of  assimilating  pigments  of  symbiotic  zooxanthellae 
in  Plexaura  honioiinilln  (Esper),  1972  forma  Kukenthali  Moser,  1921  at  various  depths. 
Proc.  U.N.  Symp.  Coop.  Im'cst  Carih.  (CICAR-II),  Caracas,  1976:  in  press. 

HAM  BERG,  M.,  AND  B.  SAMUELSSON,  1973.  Detection  and  isolation  of  an  endoperoxide  inter- 
mediate in  prostaglandin  biosynthesis.  Proc.  Natl.  Acad.  Sci.  U.S.A.,  70  :  899-903. 

HAMBKRG,  M.,  AND  B.  SAMUELSSON,  1974.  Prostaglandin  endoperoxides.  Novel  trans- 
formations of  arachidonic  acid  in  human  platelets.  Proc.  Natl.  Acad.  Sci.  U.S.A., 
71  :  3400-3404. 

HAMBEKG,  M.  J.,  SVENSSON,  AND  B.  SAMUELSSON,  1975.  Thromboxanes  :  a  ne\v  group  of  bio- 
logically active  compounds  derived  from  prostaglandin  endoperoxides.  Proc.  Natl. 
Acad.  Sci.  U.S.A.,  72  :  2994-2998. 

HINMAN,  J.  W.,  1974.  Ecology,  harvesting,  environmental  impact  and  mariculture  potential. 
Pages  150-165  in  F.  M.  Bayer  and  A.  J.  Weinheimer,  Eds.,  Prostaglandins  from 
Plexaura  homomalla  :  ecology,  utilization-  and  conservation  of  a  major  medical  marine 
resource.  University  of  Miami  Press,  Coral  Gables,  Florida. 

HINMAN,  J.  W.,  S.  R.  ANDERSON,  AND  M.  SIMON,  1974.  Studies  on  experimental  harvesting 
and  regrovvth  of  Plexaura  homomalla  in  Grand  Cayman  waters.  Pages  39-57  in 
F.  M.  Bayer  and  A.  J.  Weinheimer,  Eds.,  Prostaglandins  from  Plexaura  homomalla  : 
ecology,  utilization  and  conservation  of  a  major  medical  marine  resource.  University 
of  Miami  Press,  Coral  Gables,  Florida. 

HYMAN,  L.  H.,  1940.  The  Invertebrates  Protozoa  throuijh  Ctenophora.  McGraw-Hill,  XY\\ 
York,  726  pp. 

JOHNSON,  M.  E.,  AND  H.  J.,  SNOOK,  1955.  Seashore  animals  of  the  Pacific  Coast.  Dover, 
New  York,  659  pp. 

JOHNSON,  R.  A.,  D.  R.  MORTON,  I.  H.  KINNER,  R.  R.  GORMAN,  J.  C.  McGuiRE,  AND  F.  F. 
SUN,  1976.  The  chemical  structure  of  prostaglandin  X  (prostacyclin).  Prosta- 
glandins, 12:  915-928. 

JORDAN,  E.,  L.  CASTANARES,  AND  R.  IBARRA,  1978.  Evaluation  of  the  population  of  Plc.raui-a 
homomalla  in  relation  to  its  commercial  use.  Proc.  U.N.  Symp.  Coop.  Invest.  Carib. 
(CICAR-II).  Caracas,  1976:  in  press. 

KARIM,  S.  M.  M.  (Ed.),  1975.  Prostaglandins  and  reproduction.  University  Park  Press, 
Baltimore,  332  pp. 

KARIM,  S.  M.  M.,  (Ed.),  1976.  Prostaglandins:  physiological,  pharmacological  and  patho- 
logical aspects.  University  Park  Press,  Baltimore,  367  pp. 

KARIM,  S.  M.  M.,  AND  B.  RAO,  1975.  General  Introduction.  Pages  1-22  in  S.  M.  M.  Karim, 
Ed.,  Prostaglandins  and  reproduction.  University  Park  Press,  Baltimore. 

KINZIE,  R.  A.,  1974.  Plexaura  homomalla:  the  biology  and  ecology  of  a  harvestable  marine 
resource.  Pages  22-38  in  F.  M.  Bayer  and  A.  J.  Weinheimer,  Eds.,  Prostaglandins 
from  Plexaura  homomalla :  ccologv,  utilization  and  conservation  of  a  major  medical 
marine  resource.  University  of  Miami  Press,  Coral  Gables,  Florida. 

LANDS,  W.  E.  M.,  AND  L.  M.  ROME,  1976.  Inhibition  of  prostaglandin  biosynthesis.  Pages 
87-138  in  S.  M.  M.  Karim,  Ed.,  Prostaglandins:  chemical  and  biochemical  aspects. 
University  Park  Press,  Baltimore. 

LETELLIER,  R.  R.,  W.  L.  SMITH,  JR.,  AND  W.  E.  M.  LANDS,  1973.  Effect  of  metal-complexing 
agents  on  the  oxygenase  activity  of  sheep  vesicular  glands.  Prostaglandins.  4  :  837-843. 

LIGHT,  R.  J.,  AND  B.  SAMUELSSON,  1972.  Identification  of  prostaglandins  in  the  gorgonian, 
Plexaura  homomalla.  Eur.  J.  Biochem.,  28:  232-240. 

LOWRY,  O.  H.,  N.  J.  ROSEBROUGH,  A.  L.  FARR,  AND  R.  J.  RANDALL,  1951.  Protein  measure- 
ment with  the  Folin  phenol  reagent.  /.  liiol.  Chem..  193  :  265-275. 

MIYAMOTO,  T.,  N.  OGINO,  S.  YAMAMOTO,  AND  O.  HAYAISHI,  1976.  Purification  of  prosta- 
glandin endoperoxide  synthetase  from  bovine  vesicular  gland  microsomes.  /.  Kiol. 
Chem.,  251  :  2629-2636." 

MORSE,  D.  E.,  H.  DUNCAN,  N.  HOOKER,  AND  A.  MORSE,  1977.  Hydrogen  peroxide  induces 
spawning  in  mollusks,  with  activation  of  prostaglandin  endoperoxide  synthetase.  Science, 
196:  298-300. 

MORSE.  1).  E.,  H.  DUNCAN.  X.  HOOKER.  AND  A.  MORSF,  1978.     An  inexpeiivive  chemical  method 


452  MORSE,  KAYNE,  TIDYMAN  AND  ANDERSON 

tin"  the  control  and  synchronous  induction  of  spawning  and  reproduction  in  molluscan 

species   important   as   protein-rich   food   resources.     Proc.    U.N.    Syuip.    Coop.    Invest. 

Carih.  (CICAR-II),  Caracas,  1976:  in  press. 
OSTARELLO,  G.  I..,  1973.     Natural  history  of  the  hydrocoral  AHopora  California!  Verrill   (1866). 

Biol.  Hull.,   145:  548-564. 
PACE-ASCIAK,   C.,   AND   L.   S.   WOI.KE,    1971.     A   novel   prostaglandin   derivative   formed   from 

arachidonic  acid  by   rat   stomach   homogenates.     Biochemistry,   10:   3657-3664. 
Roos,  P.  J.,  1971.     The  shallow-water  stony  corals  of  the  Netherlands  Antilles.     Studies  Fauna 

Curacao,  37  :    1-108. 
SALMON,  J.  A.,  AND  S.  M.  M.  KARIM,  1976.     Methods  for  analysis  of  prostaglandins.     Pages 

25-86   in   S.   M.   M.    Karim,    Ed.,   Prostaglandins:    chemical   and   biochemical   aspects. 

University   Park   Press,   Baltimore. 
SAMUELSSON,    B.,    E.    GRANSTOM,    K.    GREEN,    M.    HAMBERG,    AND    S.    HAMMARSTOM,    1975. 

Prostaglandins.    Ann.  Rev.  Biocliciu.,  44:  669-695. 
SCHNEIDER,  W.  P.,  1976.     The  chemistry  of  prostaglandins.     Pages  1-24  in  S.  M.  M.  Karim, 

Ed.,  Prostafjlandins:  chemical  and  biochemical  aspects.     University  Park  Press,  Balti- 
more. 
SCHNEIDER,  W.  P.,  R.  D.  HAMILTON,  AND  L.   E.   RHULAND,    1972.     Occurrence  of  esters  of 

(15  S) -prostaglandin  As  and  E2  in  coral.     /.  Am.  Chcm.  Soc.,  94:  2122-2123. 
SMITH,  F.  G.  W.,  1971.    Atlantic  reef  corals.    University  of  Miami  Press,  Coral  Gables,  Florida, 

164  pp. 
SMITH,    R.    L,    AND   J.    T.    CARLTON,    1975.      Light's    manual:    intertidal    invertebrates    of    the 

California  coast.     University  of  California  Press,  Los  Angeles,  405  pp. 
TAKEGUCHI,   C.,   AND   C.   J.    SIH,    1972.     A   rapid   spectrophotometric   assay   for    prostaglandin 

synthetase.     Prostaglandins,  2  :  8169-8183. 
WEINHEIMER,   A.   J.,   1974.     The  discovery  of   15-epi-PGA^  in  Plcxaura   homomalla.     Pages 

17-21   in   F.   M.   Bayer  and  A.   J.   Weinheimer,    Eds.,   Prostaglandins   from    Plexaura 

homomalla  :  ecology,  utilization  and  conservation  of  a  major  medical  marine  resource. 

University  of  Miami  Press,  Coral  Gables,  Florida. 
WEINHEIMER,  A.  J.,  AND  R.  L.  SPRAGGINS,   1969.     The  occurrence  of  two  new  prostaglandin 

derivatives    (15-epi-PGA2  and   its  acetate,   methyl   ester)    in   the   gorgonian,   Ple.raura 

homomalla.      Tetrahedron  Lett.,  59  :  5185-5188. 


Reference:   Bwl.  Hull..  154  :   453-462.    (June,  1<;7S) 


ADDITIONAL  EXPERIMENTS  ON  THE  BEHAVIOR  OF  BUDS  IN 
THE  ASCIDIAN,  APLIDIUM  MULTIPLICATUM 

MITSUAKI  NAKAUCHI  AND  KAZUO  KAWAMURA 

/  >i-f>nrti/ii-nt   of  Biology,  Faculty   of   Science,   Koc/ii   University, 
Asakura,   Koclii  7X0.  Japan 

In  the  colonial  ascidian,  Aplidium  iiniltiplicatiini,  the  strobilae  produced  in  the 
abdomen  and  postabdomen  migrate  through  the  tunic  and  approach  the  regenerating 
thorax  (their  mother  zooid)  to  form  a  common  cloacal  system  with  it  (Nakauchi, 
1966a,  Nakauchi  and  Kawamura,  1974a).  In  a  previous  paper  (Nakauchi  and 
Kawamura,  1974b),  a  series  of  experiments  were  undertaken  by  the  authors  to 
study  the  mechanism  by  which  the  buds  move  in  the  "right"  direction,  and  by 
which  the  buds  and  mother  zooid  form  a  system. 

Three  kinds  of  experiments  were  described  in  the  previous  paper :  first, 
destroying  the  mother  zooid ;  secondly,  pulling  out  the  mother  zooid ;  and  thirdly, 
pulling  out  the  mother  zooid  together  with  the  tunic  covering  it.  The  results  of 
these  experiments  suggested  the  possibility  that  a  substance  secreted  by  each 
mother  zooid  diffuses  through  the  tunic  and  attracts  the  growing  buds. 

In  order  to  confirm  the  existence  of  the  attractant  and  to  determine  the  time 
and  site  of  its  secretion,  four  additional  experiments  were  designed. 

MATERIAL  AND  METHODS 

A  colonial  ascidian,  Aplidium  multiplicatum,  was  used  (see  Nakauchi  and 
Kawamura,  1974a).  The  experiments  were  done  at  the  Usa  Marine  Biological 
Station  of  Kochi  University,  from  March  to  June,  1974,  at  a  sea  water  tempera- 
ture of  18-22°  C.  For  details  of  culture  method  and  treatment  of  colonies  prior 
to  operations,  see  Nakauchi  and  Kawamura  (1974b).  The  four  experiments 
described  in  this  report  are  numbered  consecutively  with  those  of  the  previous 
paper  (Nakauchi  and  Kawamura.  1974b;  Experiments  I,  II  and  III). 

Experiment  IV 

The  results  of  Experiment  II  and  III  suggested  that  the  attractant  is  secreted 
from  a  budding  (mother)  zooid,  and  it  remains  in  the  tunic  for  a  while  even  after  the 
removal  of  the  zooid.  Experiment  IV  was  designed  to  determine  whether  the  sub- 
stance is  secreted  only  by  budding  zooids  or  whether  it  is  also  produced  by  nonbud- 
ding  zooids.  It  is  known  in  polycitorines  (Oka  and  Watanabe,  1961)  and  in  poly- 
clinids  (Freeman,  1971)  that  the  removal  of  the  thorax  of  a  zooid  is  followed 
bv  strobilation  of  the  abdominal  region  within  one  or  two  days.  So,  in  this 

j  -' 

experiment  the  thorax  of  a  grown  zooid  (a  prospective  mother  zooid)  was  cut  off, 
and  the  behavior  of  the  experimentally-produced  buds,  which  lack  a  mother  zooid 
from  the  first,  was  followed.  If  the  substance  is  secreted  into  the  tunic  even  in 
the  nonbudding  period,  the  produced  buds  would  aggregate  near  the  place  where 

453 


454 


\l     NAK.U'CHI  AND   K.   KAWAMURA 


D(2) 


E(3) 


FIGURE  1.  Behavior  of  buds  in  Experiment  IV.  Successive  stages  (A-E)  in  a  typical 
case  are  viewed  from  the  ventral  side.  The  small  arabic  numbers  identify  individual  buds  in 
order  from  anterior  to  posterior,  and  the  figures  in  parentheses  indicate  the  number  of  days 
before  or  after  budding. 

the  thorax  had  been  located  ;  if  it  is  not,  the  grown  buds  should  not  aggregate  at 
any  definite  place.  They  might  aggregate  at  various  places  by  chance,  or  open  to 
the  exterior  independently  without  grouping.  A  total  of  IS  operations  of  this 
type  were  made. 


Experiment    I ' 

In  Experiment  IV  the  growing  buds  arranged  themselves  near  the  place 
where  the  prospective  mother  zooid  had  been  located.  Therefore,  in  Experiment  V 
the  thorax  of  a  grown  zooid  was  cut  off,  together  with  the  tunic  surrounding  it 
(Fig.  3A).  l>y  this  procedure  it  was  hoped  to  eliminate  the  attracting  influence 
of  the  prospective  mother  zooid.  A  total  of  eight  operations  of  this  type  were  made. 

Experiment   I  7 

In  normal  budding,  growing  buds  arrange  themselves  around  the  atrial  aperture 
of  mother  zooid.  It  is  plausible,  therefore,  that  the  attractant  is  most  actively 
secreted  by  the  epidermis  around  the  aperture.  So,  in  Experiment  VI  the  anterior 
tip  of  the  mother  zooid  was  removed  within  one  day  after  strobilation.  At  the 


BEHAVIOR  OF  BUDS  IN  ASCIDIANS 


455 


same    time,    all    the    Imcls    hut    one    were   also    removed,    and    the    behavior    of    the 
remaining  hud  \vas  followed.     A  total  of  nine  operations  of  this  type  were  made. 

Experiment   I'll 

The  result  of  Experiment  IV  suggested  that  the  attractant  is  secreted  not  only 
by  budding  zooids  but  also  by  nonbudding  zooids.  Consequently,  Experiment  VII 
was  designed  to  find  out  whether  a  grown  (nonbudding)  zooid  has  the  potency 
to  attract  buds  produced  by  other  zooids  in  the  same  colony.  For  the  convenience 
of  observation,  all  the  zooids  but  two  were  removed  from  the  colony.  After  the 
operation  the  two  remaining  zooids  came  together,  and  a  small  colony  consisting 
of  only  two  zooids  was  formed.  It  was  known  that  budding  in  a  colony  does 
not  occur  synchronously.  In  the  present  case,  therefore,  one  zooid  was  expected 
to  make  buds  earlier  than  the  other,  and  we  could  hope  to  study  the  attractive 
influence  of  a  nonbudding  zooid  to  buds  produced  by  the  other. 


FIGURE  2.  Behavior  of  buds  in  Experiment  IV.  Successive  stages  (A-E)  in  one  of 
the  minor  cases  are  viewed  from  the  ventral  side.  Bud  identification  numbers  and  days 
before  or  after  budding  are  indicated  as  in  Figure  1. 


456 


M.  NAKAUCHI  AM)   K.   KAWAMURA 


D(2) 


1  mm 


E(3) 


FIGURE  3.  Behavior  of  buds  in  Experiment  V.  Successive  stages  (A-E)  in  a  typical 
case  are  viewed  from  the  ventral  side.  Bud  identification  numbers  and  days  before  or  after 
budding  are  indicated  as  in  Figure  1. 

RESULTS 

Experiment  IT 

In  14  cases  out  of  18,  a  single  common  cloacal  system  was  formed  by  grown 
buds  more  or  less  near  the  place  where  the  thorax  (prospective  mother  zooid) 
had  been  located  (Fig.  1).  In  the  remaining  four  cases  growing  buds  formed 
two  groups  and  finally  made  two  systems  (Fig.  2).  Even  in  the  cases  in  which 
only  one  system  was  formed,  the  behavior  of  buds  was  somewhat  different  from 
that  in  usual  budding.  Buds  lacking  their  mother  moved  more  irregularly  than 
in  usual  budding  for  about  two  days  after  budding  (Fig.  1C-D).  As  a  rule,  the 
buds  which  had  been  originally  located  apart  from  the  removed  thorax  needed  more 
time  to  find  the  right  direction  than  those  located  near  the  thorax.  Following 
this  stage  the  growing  buds  grotiped  to  form  a  common  cloacal  system. 

Experiment   V 

In  five  out  of  eight  cases  observed,  one  common  cloacal  system  was  formed  by 
new  zooid,  while  in  the  remaining  three  cases  two  systems  were  formed.  Irrespec- 


BEHAVIOR  OF  BUDS  IN  ASCIDIANS 


457 


tire  of  the  number  of  systems  formed,  the  site  of  the  common  cloacal  aperture  did 
not  appear  to  be  influenced  by  the  location  of  the  prospective  mother  zooid  which 
had  been  removed  with  its  tunic  before  budding.  That  is,  the  systems  did  not 


20(31) 


23(34) 


Ohr(11hr) 


m 


39(50) 


FIGURE  4.  Behavior  of  an  injured  mother  zooid  and  one  remaining  hud  (Experiment  VI), 
successive  stages  (A-I).  in  ventral  view.  The  outline  of  the  original  tunic  is  omitted  in 
this  figure  only.  Time  shown  outside  parentheses  indicates  the  time  after  the  operation  in 
hours.  Time  shown  in  parentheses  indicates  the  time  after  budding  in  hours.  Abbreviations 
are :  b,  bud ;  m,  mother  zooid. 


458  M.   NAKAUCHI  AND  K.   KAWAMURA 

arise  close  to  the  cut  surface  of  the  tunic  (Fig.  3).  In  most  cases  in  which  one 
system  was  formed,  the  new  common  cloacal  aperture  formed  lateral  to  the  position 
of  the  middle  of  the  mother  zooid's  abdomen  before  strobilation  (point  "X"  on 
Fig.  3A). 

Experiment   J'l 

In  all  nine  cases  the  behavior  of  the  single  remaining  bud  looked  strange.  In 
eight  cases  the  bud  first  moved  forward,  then  turned  and  approached  the  abdomen 
or  postabdomen  of  the  mother  zooid.  This  was  followed  by  a  complicated  behavior 
of  both  bud  and  mother,  the  behavior  of  one  apparently  affecting  the  behavior  of 
the  other.  After  this,  mother  and  bud  arranged  themselves  side  by  side  and  finally 
made  a  common  cloacal  system.  In  the  remaining  exceptional  case,  the  bud 
moved  away  from  its  mother  zooid,  and  each  opened  to  the  exterior  independently. 

Figure  4  shows  one  of  the  major  cases.  The  mother  zooid  contracted  strongly 
after  the  operation  ;  its  thorax  remained  contracted  for  about  20  hours,  while  its 
new  heart  began  to  beat  faintly  about  8  hours  after  the  operation.  The  bud 
moved  toward  the  mother's  thorax  during  the  first  20  hours,  then  began  to  turn 
and  finally  pointed  in  the  opposite  direction  (Fig.  4E).  After  this  the  bud  moved 
toward  the  posterior  end  of  the  mother.  On  the  other  hand  the  mother,  which 
had  been  turning  very  slowly,  began  a  complicated  behavior  as  the  bud  came  near. 
Mother  and  bud  changed  their  position  as  if  they  were  affected  by  each  other 
(Fig.  4,  F-H),  and  they  finally  arranged  themselves  side  by  side  (Fig.  41)  and 
made  a  system. 

Experiment   J'l  I 

Figure  5 A  shows  two  zooids  left  in  the  tunic,  in  which  Zooid  A  is  making 
ten  buds.  It  was  desired  to  eliminate  the  attracting  influence  of  the  thorax  of 
Zooid  A  in  order  to  see  the  attractive  effect  of  Zooid  B  upon  the  buds  produced 
by  Zooid  A.  Thus,  the  thoracic  bud  of  Zooid  A  was  extirpated  from  the  tunic. 
Four  buds  (3,  4,  9,  and  10)  were  also  cut  off  for  the  convenience  of  the 
observation.  After  these  operations.  Bud  1  and  Bud  2  gradually  moved  toward 
the  place  where  their  mother  had  been  located,  and  then  they  began  to  form  a  new 
system  by  themselves  (Fig.  5B-D).  On  the  other  hand.  Buds  5,  6,  7,  and  8 
approached  the  thorax  of  Zooid  B  (Fig.  5B)  ;  however,  Zooid  B  made  nine  buds 
two  days  after  Zooid  A  had  budded.  Of  the  nine  buds  of  Zooid  B,  five  buds 
(1,  2,  3,  5,  and  6)  were  incorporated  into  the  system  which  was  being  formed 
by  Buds  5,  6,  7,  and  8  of  Zooid  A.  The  remaining  four  buds  were,  however, 
attracted  by  Buds  1  and  2  of  Zooid  A,  and  finally  formed  a  system  with  them. 
As  shown  in  Figure  5E,  two  systems  were  formed  in  a  colony,  each  of  which 
consisted  of  zooids  of  two  different  origins. 

DISCUSSION 

The  results  of  Experiments  IV  and  V  are  consistent  with  the  hypothesis  that 
a  substance  secreted  from  the  thoracic  region  of  the  mother  diffuses  through  the 
tunic  and  attracts  buds.  It  is  plausible  that  in  Experiment  IV  the  substance 


BEHAVIOR  OF  BUDS  IN  ASCIDlANs 


4.S9 


A(0) 


FIGURE  5.  Behavior  of  buds  in  Experiment  VII,  successive  stages  (A-E),  in  ventral  view. 
Time  shown  in  parentheses  indicates  the  time  in  days  after  budding  of  zooid  A.  The  buds 
of  Zooid  A  are  identified  by  upright  arabic  numbers,  and  those  of  Zooid  B  by  sloping  ("italic") 
numbers. 


460  M.  NAKAUCHI  AND  K.  KAWAMURA 

remaining  in  the  tunic  after  removal  of  the  mother's  thorax  attracted  the  buds. 
This  suggests  that  the  period  of  secretion  is  not  restricted  to  the  time  of 
budding.  It  is  also  likely  that  in  Experiment  V  the  removal  of  the  adjacent 
tunic,  as  well  as  the  maternal  thorax,  caused  the  elimination  of  the  substance, 
and  the  buds  became  free  of  the  influence  of  the  mother.  On  the  other  hand, 
the  buds  appear  to  have  influenced  each  other,  seeing  that  they  formed  a  system 
independent  of  the  position  of  the  "prospective  mother  zooid". 

The  result  of  Experiment  VI  is  susceptible  to  various  interpretations.  The 
behavior  of  buds  which  have  temporarily  moved  from  the  thoracic  region  of  the 
injured  mother  may  be  explained  at  least  in  two  ways.  One  is  that  the  buds 
"dislike"  the  substance  secreted  by  the  wounded  thorax,  and  they  moved  away 
from  the  vicinity  of  the  wound.  Another  possibility  is  that  the  attractant 
is  constantly  secreted  all  over  the  body  surface,  although  secretory  activity  shows 
an  anteroposterior  gradient  with  its  high  point  at  the  most  anterior  region.  When 
the  anterior  end  is  removed,  the  injury  is  accompanied  by  a  reduction  in  secretory 
activity  of  the  region,  and  the  density  of  the  attractant  in  the  posterior  region 
may  become  higher  than  that  at  the  anterior.  This  temporary  reversal  of 
the  polarity  may  last  only  until  regeneration  of  the  lost  anterior  end  is  completed. 
If  this  supposition  is  the  case,  one  would  expect  the  bud  to  move  posteriorly  at 
first,  and  then,  after  the  recovery  of  the  original  polarity,  to  move  anteriorly. 
Judging  from  the  behavior  of  buds  in  other  experiments,  the  latter  case  seems 
probable. 

In  Experiment  VII  it  was  shown  that  a  bud  is  attracted  not  only  by  its  mother 
but  also  by  another  grown  (nonbudding)  zooid  or  by  a  developing  bud  of  another 
zooid.  Because  all  the  zooids  in  a  colony  originate  from  an  oozooid,  every  zooid 
has  the  same  set  of  genes.  Consequently,  zooids  in  a  colony  appear  to  have  many 
characters  in  common.  In  other  words,  they  lack  individuality  in  many  features. 
Therefore,  an  attractant  secreted  by  one  zooid  may  naturally  attract  all  the  buds 
in  a  colony  regardless  of  their  origin.  The  behavior  of  Buds  1  and  2  of  Zooid  A 
is  understandable  if  we  presuppose  an  attractant  which  had  been  secreted  by  the 
thorax  of  Zooid  A  and  had  diffused  into  the  tunic  before  the  thorax  was  removed. 
This  result  coincides  with  that  of  Experiment  III  shown  previously  (Nakauchi 
and  Kawamura,  1974b). 

Experiments  are  being  undertaken  to  get  more  direct  evidence  of  the  existence 
of  the  attracting  substance.  The  authors  do  not  necessarily  postulate  a  substance 
which  has  evolved  specifically  for  the  purpose  of  attraction.  On  the  contrary,  it 
is  likely  that  growing  buds  are  attracted  by  some  metabolite  produced  by  their 
mother  zooid. 

Setting  aside  the  possible  existence  of  an  attracting  substance,  the  movement 
of  developing  buds  is  known  in  many  colonial  ascidians.  In  polyclinids,  most  of 
which  form  common  cloacal  systems,  Brien  (1936)  seems  to  have  been  the  first  to 
describe  the  movement  of  buds  and  its  relation  to  system  formation.  Movement 
of  buds  is  known  even  in  colonial  ascidians  which  form  no  systems  (Polycitor 
iniitahilis.  Oka  and  Usui,  1944;  Mctandrocarpa  faylori,  Abbott,  1953,  and  New- 
berry,  1965;  Archidistoina  aggrcgatitui,  Nakauchi,  1966b ;  Syniplegina  re f> tans, 
Sugimoto  and  Nakauchi  1974;  Rlttcrclla  pitlchra,  Nakauchi,  1977).  Among  these 
forms  the  buds  of  Rlttcrclla  pulchni  can  turn  as  much  as  180°.  We  suggest  that 


BEHAVIOR  OF  BUDS  IN  ASCIDIANS  461 

buds  of  colonial  ascidians  have  some  ability  to  move  through  the  tunic  and  to 
change  direction  of  movement.  Some  colonial  ascidians  may  have  exploited  this 
ability  for  the  purpose  of  system  formation.  In  Metandrocarpa  it  is  known  that 
the  common  vascular  system  in  the  tunic  plays  an  important  role  in  the  movement 
of  buds  (Newberry,  1965).  Nothing  is  known  of  the  mechanism  of  bud  move- 
ment in  polyclinid  ascidians  which  have  no  common  vascular  system.  However, 
buds  and  developed  zooids  are  sometimes  observed  to  contract  and  expand,  and 
it  is  likely  that  this  action  is  involved  in  the  movement  of  zooids.  The  fact  that 
all  the  buds  and  developed  zooids  can  move  forward  only  (Nakauchi  and  Kawa- 
mura,  1974a)  may  be  a  clue  for  solving  this  problem. 


It  is  our  pleasant  duty  to  express  our  hearty  thanks  to  all  the  members  of 
the  Usa  Marine  Biological  Station,  Kochi  University,  in  which  the  study  was 
carried  out.  We  also  thank  Professor  D.  P.  Abbott  of  the  Hopkins  Marine  Station, 
Stanford  University,  for  his  encouragement  and  critical  reading  of  the  manuscript. 

SUMMARY 

In  a  previous  paper  by  the  authors,  it  was  suggested  that  the  behavior  of 
growing  buds,  which  form  a  common  cloacal  system,  is  affected  by  a  substance  which 
is  secreted  by  the  mother  zooid  and  diffuses  through  the  tunic. 

Four  sets  of  experiments  were  made  to  confirm  the  existence  of  the  substance, 
and  to  get  more  information  about  the  attractant.  In  the  first  set,  the  thorax 
of  a  grown  zooid  was  removed  before  budding,  and  artificial  strobilation  was  in- 
duced. In  this  case  the  buds  lacked  the  mother  zooid  from  the  first.  In  the 
second,  the  thorax  of  a  grown  zooid  was  removed  before  budding,  together  with 
the  tunic  covering  the  thorax.  In  the  third,  the  anterior  tip  of  a  mother  zooid, 
thought  to  be  a  center  of  secretion,  was  removed.  In  the  fourth,  the  experiment 
was  designed  to  show  whether  a  bud  is  attracted  only  by  its  mother  and  sisters 
or  also  by  other  zooids  in  the  same  colony.  After  these  operations  the  behavior 
of  buds  was  followed. 

The  results  supported  the  existence  of  the  attractant.  They  suggested  that 
the  time  of  secretion  is  not  restricted  to  the  period  of  budding,  that  the  site  of 
secretion  is  not  restricted  to  a  special  region  of  the  zooid,  and  that  a  bud  is 
attracted  not  only  by  its  mother  but  also  by  any  other  zooid  in  the  same  colony. 

LITERATURE  CITED 

ABBOTT,   D.    P.,   1953.     Asexual   reproduction   in   the   colonial   ascidian    Metandrocarpa    tayluri 

Huntsman.     Univ.  Calif.  Pitbl.  Zoo!.,  61  :  1-78. 
BRIEN,    P.,    1936.      Formation   des    coenobies    chez    les    Polyclinidae :    Circinalium    concresccns 

(Giard)  =  Sidnvuin   turbinatum    (Savigny),  var.   concrcsccns    (Giard).     .  Inn.  Soc.  R. 

Zool.  Bclg.,  67  :~63-73. 
FREEMAN,  G.,   1971.     A  study  of  the   intrinsic  factors   which   control   the   initiation  of  asexual 

reproduction  in  the  tunicate  Amaroucium  constellatum.     J.  E.rp.  Zool.,   178:  433-456. 
NAKAUCHI,   M.,   1966a.     Budding  and  colony  formation  in  the  ascidian,   Amaroucium    nnilti- 

plicatum.     Jpu.  J.  Zool.,  15:   151-172. 
NAKAUCHI,  M.,  1966b.     Budding  and  growth  in  the  ascidian,  Arclndistonui  iii/t/rct/tituiii.     Rep. 

Usa  Mar.  Biol  Stn.,  13  :  1-10. 


462  M.    NAKAUCIM    AND   K.   KAWAML'kA 

NAKAUCHI,    M.,    19/7.      Development    and    budding    in    tin-    oozooid    of    polyclinid    ascidians. 

2.     Ritterelhi  pulclini.   Annot.  /<><>!.  .//>»..  50:   151-159. 
\'.\K.\i'(  HI,    M.,   AND   K.    KANVAMTKA,    1974a.      The   behavior   of   buds   during    common   cloacal 

system  formation  in  the  ascidian,  Aplidimn   multiplication.     Rep.    Csa   Mur.  ]1iol.  Stn., 

21  :   19-27. 
NAKAUCHI,  M.,  AND  K.   KAWAMURA,   1974b.     Experimental  analysis  of  the  behavior  of  buds 

in  tlie  ascidian,  Aplidiuni  multiplicatum.     I.    Rep.  Lisa  Mur.  liiol.  Stii.,  21  :  29-3<S. 
NEWBERRY,  A.   T.,   1965.     The  structure  of  the  circulatory  apparatus  of  the  test  and   its   role 

in    budding    in    the    polystyelid    ascidian,    Metandrocarpa    tavlori    Huntsman.       A'lcin. 

Acatl.  R.  Belg.  Cl.  Sci.  Scr.  4°,  16:  1-57. 
OKA,   H.,   AND  M.   Usui,    1944.     On  the  growth   and  propagation  of  the  colonies   in  Polycitor 

i/iiitdl'ilis    (Ascidiae    compositae).      Sci.    Rep.    Tak\o    liuurika    nuii/uku,    Sect.    li,    1  : 

23-53. 
OKA,  H.,  AND  H.  WATANABE,  1961.     Kimstriche  Auslosung  der  Strobilation  bei  den  Synascidien. 

Emhryoloi/in.   6:    135-150. 
SUGIMOTO,  K.,  AND  M.   NAKAUCHI,   1974.     Budding,   sexual   reproduction,  and  degeneration   in 

the  colonial  ascidian,  Syiiiplc;/iiKi  reptans.     Hiol.   Hull..   147  :   213-226. 


Reference:  Bwl.  Hull..  154:  -lfo-471.      (June,   1978) 


SEASONAL  BURROWING  BEHAVIOR  AND  ECOLOGY 

OF  APORRHAIS  OCCIDENTALS 

(GASTROPODA:  STROMBACEA) 

FRANK  E.  PERRON  ' 
Department   of  Zoology.   University  of  Nciv  Hampshire,   !>iirhn/n,   Ne?c   Hampshire 

The  mesogastropod  family  Aporrhaidae  is  represented  by  only  three  living 
species  comprising  the  genus  Aporrhais.  Aporrlmls  pcspclccani  (L. )  and  A. 
serresiana  (Michaud)  are  restricted  to  the  eastern  Atlantic  and  have  been  studied 
by  Yonge  (1937).  [See  Eretter  and  Graham  (1962)  for  a  general  account  of 
the  natural  history  of  these  species.]  Locomotion  in  A.  pcspclccani  has  been 
examined  by  Weber  (1925)  and  by  Haefelfinger  (1968).  Aporrhais  occidcntalis 
(Beck)  ranges  from  Labrador  to  Massachusetts  in  the  western  Atlantic  (Johnson, 
1930)  and  is  found  in  depths  of  water  from  10-2000  m  (Clarke,  1962).  Little 
information  is  available  on  A.  occidentalis. 

Aporrhaids  are  of  particular  interest  to  malacologists  because  they  are  the 
most  primitive  members  of  the  superfamily  Strombacea  which  includes  the  widely 
distributed  and  conspicuous  tropical  genera  S  trombus  and  Lam  bis.  According 
to  Cox  (1960)  and  Zittcll  (1913),  aporrhaids  first  appeared  in  the  Jurassic  as 
the  earliest  representatives  of  the  Strombacea,  and,  on  the  basis  of  shell  structure, 
Morton  (1956)  considers  A.  occidcntalis  to  he  the  most  primitive  living  aporrhaid. 
As  is  typical  of  most  members  of  the  Strombacea,  the  shell  of  Aporrhais  is  subject 
to  age-dependent  changes  in  morphology.  The  expanded  and  thickened  outer 
shell  lip  of  adults  is  absent  in  juveniles. 

The  Aporrhaidae,  as  well  as  the  related  but  less  ancient  Struthiolariidae 
(Morton,  1951),  are  known  to  burrow  in  soft  marine  sediments,  and  Schafer 
(1972)  has  commented  on  the  importance  of  Aporrhais  in  reworking  the  substrate. 
Yonge  (1937)  described  the  burrowing  behavior  of  A.  pcspclccani  and  A.  ser- 
resiana under  laboratory  conditions  and  concluded  that  these  gastropods  are 
specialized  for  burrowing  in  muddy  gravel  and  only  rarely  move  about  on  the 
surface  of  the  substrate.  Barnes  and  Bagenal  (1952)  examined  dredged  specimens 
of  both  species  and  found  that  the  shells  of  adult  snails  were  frequently  encrusted 
with  barnacles,  bryozoans  and  polychaete  tubes.  Based  on  this  evidence,  they 
suggested  that  Aporrhais  spends  more  time  on  the  surface  of  the  mud  than  was 
previously  thought. 

The  SCUKA  techniques  used  in  the  present  study  of  A.  occidcntalis  have  per- 
mitted in  situ  tagging  experiments  and  observations  on  the  burrowing  of  these 
gastropods  in  their  natural  habitat. 

1  Present  address:  Department  of  Zoology,  Edmondson  Hall,  University  of  Hawaii, 
Honolulu,  Hawaii  96822. 

463 


464  FRANK  E.  PERRON 

MATERIALS  AND  METHODS 

During  1973-1976  a  population  of  A.  occidental-is  was  studied  in  17  m  of  water 
at  the  Isles  of  Shoals  off  Portsmouth,  New  Hampshire  (42°  59'  N,  70°  37'  W). 
The  size  structure  and  density  of  this  population  was  determined  through  quantita- 
tive bottom  sampling  using  SCUBA  transects  and  an  epibenthic  sled  (Hessler  and 
Sanders.  1967). 

In  April.  1975,  individual  snails  were  tagged  so  that  their  movements  both 
upon  and  within  the  substrate  could  be  followed  from  month  to  month.  Nylon 
fishing  line  was  used  to  affix  numbered  plastic  tags  to  the  shell  spires  of  20  male 
and  20  female  specimens  of  A.  occidentals.  The  highly  visible  tags  were  buoyant 
and  floated  5-8  cm  above  the  mud  at  all  times.  The  tagged  animals  were  placed 
around  a  cinder-block  anchor  from  which  10  m  transect  lines  were  extended  in 
the  four  compass  directions.  The  transect  lines  were  marked  at  1  m  intervals  so 
that  the  snails  could  be  located  within  the  resulting  grid  system.  From  May, 
1975,  to  May,  1976,  monthly  SCUBA  dives  were  made  on  this  site.  In  addition  to 
daytime  observations,  night  dives  were  made  in  summer  and  winter.  Data  were 
taken  on  the  location  of  tagged  snails  within  the  grid  system  and  on  whether  or 
not  these  animals  were  epifaunal  or  infaunal. 

During  each  monthly  dive,  bottom  water  temperatures  were  recorded  with  a 
hand-held  mercury  thermometer  and  notes  were  taken  on  the  occurrence  of 
potential  predators  within  the  transect  area.  Specimens  of  A.  occidcntalis  were 
collected  each  month  and  preserved  for  subsequent  gut  content  analyses.  Un- 
tagged  animals  were  normally  used  for  this  purpose,  tagged  snails  being  sacrificed 
only  when  no  others  could  be  found.  Empty  A.  occidcntalis  shells  brought  up 
in  dredge  hauls  or  found  during  SCUBA  dives  were  examined  for  evidence  of 
predation. 

In  the  laboratory,  adults  and  juveniles  of  A.  occidcntalis  were  maintained  in  a 
flowing  seawater  system.  Burrowing,  feeding  and  copulation  were  observed  and 
attempts  were  made  to  determine  the  effects  of  different  water  temperatures  on 
burrowing  behavior. 

RESULTS 

Specimens  of  A.  occidcntalis  were  first  observed  by  the  author  at  the  Isles  of 
Shoals  during  a  SCUBA  dive  in  March,  1973.  The  animals  were  fully  exposed  on 
the  level  muddy  bottom  and  seemed  to  be  grazing  on  a  thin  brown  film  which 
covered  the  substrate.  This  film  was  later  examined  and  found  to  consist  of 
high  concentrations  of  the  benthic  diatom  Pleurosigma  sp.,  as  well  as  the  decaying 
remains  of  several  species  of  macroalgae.  Gut  content  analyses  revealed  that  this 
material  was  indeed  being  ingested  along  with  some  sand,  sponge  spicules  and 
empty  foraminifera  tests.  The  shells  of  these  snails  were  not  encrusted  with 
sessile  organisms  except  that  the  shells  of  older  specimens  of  A.  occidcntalis  were 
frequently  riddled  by  the  boring  spionid  polychaete  Polydora  coinincnsalis  Andrews. 

A  series  of  thirty  1  m  X  15  m  SCUBA  transects  run  at  the  study  site  in  April, 
1974,  yielded  a  total  of  28  epifaunal  specimens  of  A.  occidcntalis.  Twenty-one 
of  these  animals  were  mature  adults  with  well-developed  outer  shell  lips,  while 
the  remaining  seven  were  juveniles  ranging  in  shell  length  from  20-45  mm.  Epi- 


SEASONAL  BURROWING  IN  APORRHAIS  465 

benthic  sled  hauls  taken  in  the  same  area  contained  large  numbers  of  juvenile 
A.  occidcntalis  not  seen  during  the  SCUBA  transects.  Ten  0.5  m  X  15  m  sled 
hauls  yielded  40  young  snails  and  only  four  adults.  Therefore,  most  of  the 
juveniles  in  this  population  were  infaunal,  while  the  adults  were  epifaunal. 
Laboratory  observations  over  a  three  year  period  also  showed  that  juveniles 
burrow  more  rapidly  and  spend  more  time  in  the  substrate  than  do  adults. 
Sediment  samples  taken  in  April,  1975,  contained  early  post-metamorphic  A. 
occidcntalis  juveniles  measuring  1.2-1.5  mm  in  shell  length.  Similar  sediment 
samples  taken  in  October,  1975,  contained  no  juveniles  smaller  than  6.5  mm. 

In  both  1973  and  1974  the  population  of  A.  occidcntalis  at  the  Isles  of  Shoals 
disappeared  from  the  surface  of  the  mud  by  August  and  did  not  reappear  until  the 
following  February.  Although  dredging  carried  out  during  the  winter  of  1974— 
1975  showed  that  the  snails  had  burrowed  at  the  study  site,  tagging  experiments 
begun  in  April,  1975,  provided  more  detailed  and  quantitative  data  on  seasonal 
burrowing  behavior. 

Figure  1  shows  the  percentages  of  tagged  specimens  of  A.  occidentalis  found 
burrowing  each  month  from  May,  1975,  through  April,  1976.  Figure  1  also 
includes  monthly  bottom  water  temperatures.  Numbers  of  burrowing  animals  are 
expressed  as  percentages  because  the  total  number  of  snails  found  each  month 
(both  infaunal  and  epifaunal)  varied  as  a  function  of  water  clarity  and  the  time 
available  for  searching.  Also,  the  number  of  tagged  snails  diminished  over  time 
as  animals  were  sacrificed  for  gut  content  analyses  or  were  lost  due  to  predation 
or  other  factors. 

Virtually  all  of  the  tagged  animals  were  infaunal  from  August  through  October. 
In  November,  all  of  the  males  remained  infaunal  but  eight  of  the  ten  females 
counted  were  epifaunal.  In  December  and  January  the  entire  population  was  again 
infaunal.  Most  of  the  tagged  A.  occidentalis  were  found  crawling  about  on  the 
surface  of  the  substrate  from  February  through  June,  and  40%  were  epifaunal 
in  July.  Except  during  the  month  of  November,  there  were  no  obvious  differences 
in  burrowing  behavior  between  male  and  female  snails. 

The  results  of  gut  content  analyses  performed  on  specimens  of  A.  occidcntalis 
collected  at  the  study  site  suggest  seasonal  changes  in  feeding  behavior  correlated 
with  burrowing.  From  August  through  January,  all  animals  had  empty  stomachs 
and  intestines.  Epifaunal  snails  collected  from  February  through  July  were 
actively  feeding  and  had  full  guts.  Furthermore,  each  of  these  animals  had  a 
well-developed  crystalline  style  in  its  style  sac.  Crystalline  styles  were  never 
found  in  animals  with  empty  guts. 

Gut  content  analyses  were  also  performed  on  specimens  of  A.  occidcntalis 
collected  in  deep  water  by  the  United  States  National  Marine  Fisheries  Service 
and  made  available  by  Dr.  Roland  Wigley  of  the  Woods  Hole,  Massachusetts, 
office  of  the  NMFS.  Three  specimens  (two  females  and  one  male)  dredged  from 
174  m  (42°  05'  N,  69°  50'  W)  in  November,  1958,  had  empty  guts.  Six  speci- 
mens (three  males  and  three  females)  collected  from  242  m  (43°  19'  N,  67°  45' 
W)  in  June,  1961,  had  full  guts. 

Because  field  observations  were  made  only  at  monthly  intervals,  it  was  impos- 
sible to  obtain  detailed  data  on  the  mobility  of  epifaunal  snails.  From  May 
through  July,  1975,  when  most  specimens  of  A.  occidental!*  were  actively  feed- 


466 


I  RANK  E.  PEKKOX 


-I  15 


100  r 


Mar.      Apr. 


FIGURE  1.  Percentages  of  tagged  A.  occitlcutalis  found  burrowing  each  month  at  the 
Isles  of  Shoals  study  site.  Numbers  over  histogram  bars  indicate  the  total  number  of  tagged 
snails  counted  each  month.  Bottom  water  temperatures  are  represented  by  connected  dots. 
Temperatures  represent  single  measurements  taken  during  monthly  dives. 

ing  on  the  surface  of  the  substrate,  no  animal  was  observed  to  move  more  than 
10  m  from  one  month  to  the  next.  When  the  population  of  tagged  A.  occidentalis 
became  infaunal  in  August,  29  of  the  original  40  snails  were  still  within  the  limits 
ol  the  transect  lines.  The  11  animals  not  counted  in  August  may  have  wandered 


SEASONAL  BURROWING  IN  APORRHAIS 


467 


away  from  the  study  area.  However,  it  is  also  possible  that  they  were  carried  off 
by  predators  or  had  lost  their  numbered  tags.  From  August  through  January, 
with  the  exception  of  November,  no  movements  of  individual  snails  were  noted  from 
month  to  month.  Observations  made  during  night  dives  showed  that  although 
A.  occidental  is  is  more  active  at  night  than  in  the  daytime  during  its  epifaunal 
period  (February-July),  burrowed  snails  during  August-January  do  not  emerge 
from  the  substrate  at  night. 

Although  copulation  was  never  observed  in  the  field,  specimens  of  A.  acci- 
dent alls  kept  in  the  laboratory  frequently  copulated  at  night  during  March  and 
April. 

Potential  predators  on  A.  occidentalis  include  the  carnivorous  gastropod  Colus 
stimpsoni  Morch,  the  crab  Cancer  irroratus  Say,  and  possibly  the  molluscivorous 
wolf  fish  Anarhichas  lupus  L.  Coins  stimpsoni  is  present  at  the  Isles  of  Shoals 
study  site  throughout  the  year  and  preys  on  a  variety  of  gastropods.  Although 
C.  stimpsoni  was  not  observed  actually  feeding  on  A.  occidentals  in  the  field, 
instances  of  predation  did  take  place  in  the  laboratory.  Aporrhais  occidental  is 
shows  a  distinct  escape  response  (accelerated  locomotion)  to  the  presence  of  C. 
stimpsoni  (Perron,  1978).  The  crab  C.  irroratus  was  active  at  the  study  site 
from  July  through  November.  Several  instances  of  attempted  predation  on  A. 
occidentalis  were  observed  in  the  field,  and  in  one  case,  a  crab  was  seen  grasping 
the  numbered  tag  of  a  burrowed  A.  occidentalis  and  pulling  the  snail  from  the 
substrate.  In  the  laboratory,  crabs  readily  devoured  juvenile  A.  occidentalis  by 
progressively  cracking  away  the  shell  aperture  until  the  soft  parts  were  exposed. 
However,  even  large  specimens  of  C.  irroratus  (carapace  width  6  cm)  were  rarely 
able  to  feed  on  an  adult  A.  occidentalis  with  well-developed  outer  shell  lips. 

In  Table  I  the  71  empty  A.  occidentalis  shells  collected  haphazardly  over  the 
course  of  a  year  at  the  Isles  of  Shoals  are  classified  according  to  types  of  visible 
shell  damage.  Shells  showing  crab  damage  all  had  apertures  which  were  chipped 
away  in  the  manner  observed  in  the  laboratory  and  as  described  and  figured  by 
Vermeij  (1976).  Five  of  the  adult  A.  occidentalis  shells  showing  crab  damage 
had  previously  been  weakened  by  infestations  of  the  boring  polychaete  Polydora 
coimnensalis.  Shells  so  badly  crushed  that  they  were  reduced  to  fragments  may 
have  been  attacked  by  fish  or  crabs.  Finally,  undamaged  empty  shells  may  indicate 
predation  by  Coins  stimpsoni  or  some  undetermined  cause  of  mortality. 

Laboratory  attempts  to  influence  the  seasonal  burrowing  behavior  of  A.  occi- 


TAULE  I 

The  condition  of  empty  A.  occidentalis  shells  collected  over  a  one  year  period  at  the  Isles  of  Shoals 
study  site.     During  the  same  period  143  live  animals  (60  adults  and  83  juveniles)  were  found. 


Type  of  shell  damage 

Number  of  shells 

Probable  predator 

Adult 

Juvenile 

Chipped  outer  lip 
Crushed 

6 
2 

41 
4 

Crab 
Fish  or  crab 

No  damage 

11 

20 

Predatory  gastropod 

468  FRANK  E.  PERRON 

il CH tails  by  manipulating  water  temperature  were  unsuccessful.  Twenty  active 
epifaunal  adult  snails  collected  in  March,  1976,  were  split  into  two  groups  and 
kept  at  water  temperatures  of  4-7°  C  and  13-16°  C,  respectively.  No  differences 
in  behavior  were  noted  between  the  two  groups,  and  all  20  animals  remained 
epifaunal  until  the  experiment  was  terminated  after  two  months.  Specimens  of 
A.  Occident  a!  is  kept  in  the  laboratory  for  long  periods  of  time  tended  to  become 
less  active  and  more  infaunal.  Such  animals  were  also  subjected  to  differing  tem- 
perature regimes,  but  no  resultant  changes  in  burrowing  behavior  were  observed. 

DISCUSSION 

The  results  of  the  experiments  reported  here  show  that  the  specimens  of 
A.  occidentalis  in  the  population  studied  alternate  between  periods  of  epifaunal 
feeding  activity  and  infaunal  nonfeeding  quiescence.  Although  tagging  data  are 
available  only  for  the  year  1975-1976,  SCUBA  observations  during  the  preceding 
two  years  indicate  that  seasonal  burrowing  is  a  regular  occurrence  in  this  gas- 
tropod. Since  A.  occidentalis  has  such  an  extensive  bathymetric  range,  it  may  not 
be  reasonable  to  assume  that  the  shallow  water  Isles  of  Shoals  population  is 
typical  of  the  species  as  a  whole.  However,  gut  content  data  from  specimens 
collected  in  deeper  water  (174-242  m)  conform  precisely  to  the  pattern  observed  in 
the  Isles  of  Shoals  population. 

The  observations  of  Barnes  and  Bagenal  (1952)  on  dredged  A.  pespelecani 
are  consistent  with  a  seasonal  burrowing  pattern  similar  to  that  of  A.  occidentalis. 
The  shells  of  A.  pespelecani  collected  by  Barnes  and  Bagenal  in  April  were  covered 
with  small  newly  set  barnacles,  while  "enormously  elongated"  barnacles  were  found 
on  specimens  dredged  in  late  July.  The  presence  of  live  barnacles  indicates  that 
these  A.  pespelecani  were  epifaunal  during  the  spring  and  summer  months. 
Barnes  and  Bagenal  also  reported  that  the  shells  of  dredged  juvenile  A.  pespelecani 
were  nearly  always  free  of  encrusting  organisms.  Their  suggestion  that  juveniles 
spend  more  time  burrowed  than  do  adults  is  supported  by  the  field  and  laboratory 
observations  in  the  present  study. 

The  data  in  Figure  1  suggest  a  possible  relationship  between  water  tempera- 
ture and  burrowing.  At  the  Isles  of  Shoals  study  site,  specimens  of  A.  occi- 
dentalis emerge  from  the  substrate  when  water  temperatures  are  at  their  lowest, 
and  remain  active  until  warming  takes  place  during  the  summer.  However, 
laboratory  experiments  failed  to  provide  evidence  for  a  causal  relationship  between 
temperature  and  burrowing.  Furthermore,  since  A.  occidentalis  ranges  to  a  depth 
of  2000  m  where  seasonal  temperature  fluctuations  are  small  (Rokop,  1974), 
temperature  would  seem  an  unlikely  coordinator  of  seasonal  burrowing.  Further 
research  will  be  necessary  to  identify  the  environmental  factor  or  factors  which 
control  burrowing  in  A.  occidentalis. 

Similarly,  the  available  data  are  not  sufficient  to  explain  the  role  of  seasonal 
burrowing  in  the  life  history  of  this  gastropod.  It  is  tempting  to  suggest  that 
A.  occidentalis  avoids  predation  by  C.  irroratus  by  burrowing  at  the  time  of  year 
when  the  crab  is  most  active.  Jeffries  (1966)  has  shown  that  the  temperature 
optimum  of  C.  irroratus  is  approximately  14°  C  and  that  these  predators  become 
less  active  and  move  to  deeper  water  during  cold  winter  months.  Nevertheless, 


SEASONAL  BURROWING  IN  APORRHAIS  469 

this  explanation  for  the  seasonal  burrowing  of  A.  occidentals  seems  questionable 
when  one  again  considers  that  the  Isles  of  Shoals  population  is  at  the  shallow 
water  end  of  a  bathymetric  range  which  extends  into  the  thermally  stable  depths 
where  predators  are  presumably  not  affected  by  seasonal  temperature  fluctuations. 

Although  little  is  known  about  reproduction  in  A.  occidentalis,  Johansson 
(1948)  has  studied  the  reproductive  system  of  A.  pespclecani,  while  Lebour 
(1933)  has  observed  the  eggs  and  larvae  of  this  eastern  Atlantic  species.  Aporr- 
hais  ^cspclccani  eggs  are  small  (0.25  mm)  and  are  deposited  singly  or  in  small 
groups.  The  larvae  are  planktotrophic  and  undergo  considerable  growth  in  the 
plankton  before  settling  (Lebour,  1933;  Thorson,  1946). 

The  eggs  and  larvae  of  A.  occidentalis  have  never  been  reported.  However, 
young  benthic  animals  with  shells  measuring  1.2-1.5  mm  collected  by  the  author 
in  April,  1975,  were  nearly  identical  to  early  post-metamorphic  juveniles  of 
A.  pespclecani  figured  by  Lebour  (1933).  Since  no  juveniles  smaller  than  6.5 
mm  were  taken  in  October,  1975,  it  is  possible  that  the  breeding  season  of 
A.  occidentalis  is  similar  to  that  of  A.  pespclecani,  with  egg  laying  taking  place  in 
early  spring  (February-March)  and  larvae  settling  in  April  and  May  (Lebour, 
1933).  If  this  is  the  case,  then  the  reproductive  cycle  of  A.  occidentalis  may 
consist  of  a  build  up  of  energy  reserves  during  the  epifaunal  feeding  period 
followed  by  conversion  of  this  energy  into  gonad  development  during  the  period 
of  infaunal  quiescence.  The  presence  of  epifaunal  nonfeeding  females  in  No- 
vember is  perplexing  and  may  indicate  that  oviposition  takes  place  at  this  time 
rather  than  in  the  spring. 

Aporrliais  is  not  unique  among  the  Strombacea  in  possessing  burrowing 
habits.  The  Struthiolariidae,  which  probably  evolved  directly  from  the  Apor- 
rhaidae  (Morton,  1951),  remain  infaunal  for  long  periods  while  feeding  by  a 
ciliary  mechanism  similar  to  that  of  the  burrowing,  nonstrombacean  mesogas- 
tropocl,  Turritclla  (Yonge,  1946).  However,  there  is  no  suggestion  in  the  litera- 
ture that  burrowing  in  the  Struthiolariidae  is  seasonal,  and  unlike  Aporrliais, 
these  gastropods  certainly  continue  feeding  while  burrowed.  The  strombid 
Terebelliini  terebelluin  (L.)  is  also  known  to  be  an  active  burrower  (Abbott, 
1962),  but,  again,  year  round  studies  have  not  been  carried  out. 

A  seasonal  burrowing  cycle  similar  to  that  of  AporrJiais  has  been  described 
for  the  tropical  strombid  gastropod  Stroinbus  piigilis  L.  by  Percharde  (1968, 
1970).  Percharde  reports  that  colonies  of  S.  piigilis  off  the  island  of  Trinidad 
in  the  Caribbean  burrow  in  November,  cease  feeding,  and  do  not  resume  normal 
activity  until  March  or  April.  At  the  end  of  this  infaunal  period  the  males  emerge 
from  the  substrate  first,  while  the  females  remain  burrowed  for  a  time  and  lay 
their  eggs.  Percharde  (1970)  also  presents  data  suggesting  similar  burrowing 
behavior  in  S.  alatits  Gmelin  and  5".  raninus  Gmelin. 

Recent  studies  by  Berg  (1974)  and  Perron  (1978)  have  pointed  out  the 
marked  homogeneity  of  locomotory  behavior  patterns  within  the  Strombacea 
from  the  primitive  Aporrliais  to  the  more  highly  evolved  Stroinbus  and  Lamb  is. 
Until  year  round  in  situ  studies  have  been  carried  out  on  additional  members  of 
the  Strombacea,  it  will  not  be  possible  to  determine  how  pervasive  the  trend 
toward  seasonal  burrowing  may  be  within  this  superfamily.  Nevertheless,  the 
similarities  in  burrowing  habits  between  A.  occidentalis  and  S.  piigilis  probably 


470  FRANK  E.  PERRON 

rctUvi   the  conservative  nature  of  behavioral  evolution  within  the  morphologically 
diverse  Strombacea. 


I  thank  Larry  Harris  and  Ruth  Turner  for  their  encouragement  and  support 
during  this  project.  Technical  assistance  was  provided  by  Cynthia  Mroch.  Also, 
much  of  the  field  work  could  not  have  been  done  without  the  diving  assistance 
of  Brian  Kivest,  Barry  Spracklin,  Alan  Hulburt  and  Paul  Lavoie.  Special 
thanks  are  due  Ned  Mclntosh,  captain  of  the  University  of  New  Hampshire 
research  vessel,  JERE  A.  CHASE. 

SUMMARY 

1.  SCUBA  observations  and  in  situ  tagging  experiments  were  carried  out  on  a 
population    of   Apurrluris   occid entails   during    1973-1976.      Seasonal    changes    in 
burrowing   behavior   were    quantified    by    determining    the    percentage    of    tagged 
snails  found   burrowing  each   month.      Gut  content   analyses   were   performed   at 
monthly  intervals  to  determine  if  the  intensity  of  feeding  activity  fluctuates  sea- 
sonally.    Empty  A.  occid  entails  shells  were  collected  and  examined  for  evidence 
of  predation. 

2.  Specimens  of  A.  occidental  is  alternate  between  periods  of  epifaunal  activity 
and  infaunal  quiescence.     Tagged  snails  tended  to  remain  burrowed  from  August 
through  January,  but  were  active  on  the  surface  of  the  substrate  from  February 
until   late    summer.      Gut   content   analyses    showed    that   the    snails    fed    actively 
during  their  epifaunal  period,  but  ceased  feeding  while  burrowed. 

3.  Laboratory    attempts    to    influence    burrowing    behavior    by    manipulating 
water  temperature  were  unsuccessful. 

4.  Published   observations   on   eastern   Atlantic    species   of   Aporrhals   suggest 
that  seasonal  burrowing  behavior  may  be  characteristic  of  the  genus. 

LITERATURE  CITED 

ABBOTT,  D.  P.,  1962.     Observations  on  the  gastropod  Tcrcbctlmn  tcrcbeUitin    (Linnaeus)  ;   with 

particular  reference  to  the  behavior  of  the  eyes  during  burrowing.      Vcliuer,  5  :    1-3. 
BARNES,  H.,  AND  T.  B.  BAGENAL,  1952.     The  habits  and  habitat  of  Aporrhais  pespclicani  (L.). 

Proc.  Mahicol.  Soc.  Loud.,  29  :  101-105. 
BERG,  C.  J.,  JR.,  1974.     A  comparative  ethological  study  of  strombid  gastropods.     Behavior,  51  : 

274-322. 
CLARKE,  A.,  1962.    Annotated  list  and  bibliography  of  the  abyssal  molluscs  of  the  world.     Niitl. 

Mus.  Can.  Bull..  181  :  20. 
Cox,  L.  R.,  1960.    Thoughts  on  the  classification  of  the  gastropoda.    Proc.  Mahicol.  Soc.  Loud.. 

33:  239-261. 

FRETTER,  V.,  AND  A.  GRAHAM,  1962.    British  prosobnnicli  molluscs.    Ray  Society,  London. 
HAEFELFINGER,  R.,   1968.     Lokomotion  von  Aporrhais  pes-pelicani.     Rei'iie  Siiisse  Zool.,  75  : 

569-574. 
HESSLER,  R.  R.,  AND  H.  L.  SANDERS,  1967.     Faunal  diversity  in  the  deep  sea.     Deep  Sea  Res., 

14:  65-78. 
JEFFRIES,  H.   P.,   1966.     Partitioning  of  the  estuarine  environment   by  two   species   of   Cancer. 

Ecology,  47:   477-481. 

JOHANSSON,   J.,    1948.     liber   die    Geschlechtsorgane   von   Aporrhais   pespelecani   nebst   einigen 
.  Betrachtungen  iiber  die  phylogenetische   Bedeutung  der   Cerithiacea   und   Architaenio- 

glossa.    Arkiv.  Zool..  41 A  :  1-13. 


SEASONAL  BURROWING  IN  APORRHAIS  471 

JOHNSON,  C.  W.,  1930.     The  variations  of  Aporrhais  occidentalis   (Beck).     Nautilus,  44:   1-4. 
LEBOUR,  M.  V.,   1933.     The  eggs  and  larvae  of  Turitella  conununis  Lam.  and  .-Iporrhais  pcs- 

pelicani  (L.).    /.  Mar.  Biol.  Assoc.  U.K.,  18  :  499-51  If.. 
MORTON,  J.  E.,  1951.     The  ecology  and  digestive  system  of  the  Struthiolariidae   (Gastropoda). 

Q.  J.  Microsc.  Sci.,  92  :  1-25. 
MORTON,  J.  E.,  1956.     The  evolution  of  Perissodonta  and  Tylnspira  (Struthiolariidae).     Trans. 

R.  Soc.  N.  Z..83:  515-524. 
PERCHARDE,  P.  L.,  1968.     Notes  on  distrihution  and  underwater  ohservations  on  the  molluscan 

genus  Stroinlnis  as  found  in  the  waters  of  Trinidad  and  Tohago.     Carib.  J.  Sci.,  8  : 

47-55. 
PERCHARDE,  P.  L.,   1970.     Further  underwater  observations  on  the  molluscan  genus  Stromlnts 

Linne  as  found  in  the  waters  of  Trinidad  and  Tohago.     Carib.  J.  Sci..  10  :  73-77. 
PERRON,  F.  E.,  1978.     Locomotion  and  shell  righting  behaviour  in  adult  and  juvenile  Aporrhais 

occidentalis  (Gastropoda:  Strombacea).     Anim.  Behav.,  in  press. 

ROKOP,  F.  J.,  1974.     Reproductive  patterns  in  the  deep-sea  benthos.     Science,  186:  743-745. 
SCHAFER,  W.,   1972.     Ecolof/y  and  palacoccoloijy  of  marine  environments.     University  of  Chi- 
cago  Press,   Chicago. 

THORSON,   G.,   1946.     Reproduction  and  larval  development  of  Danish   marine  bottom   inverte- 
brates  with    special    reference    to   the    planktonic    larvae    in    the    Sound.      (Oresund). 

Mcdd.  Koiiun.  Damn.  Fiskcri0g  Haininders.     Ser.  Plankton,  4:  1-523. 
VERMEIJ,  G.  T.,  1976.     Interoceanic  differences  in  vulnerability  of  shelled  prey  to  crab  predation. 

Nature.  260  :  135-136. 
WEBER,   H.,   1925.     Uber  arhythmische   Fortbewegung  bei  einigen   Prosobranchiern,   Z.    Vcrql. 

Physiol.,  2:   109-121. 
VONGE,  C.  M.,  1937.     The  biology  of  Aporrhais  pcs-pelicani   (L.)   and  A.  scrresiana    (Mich.). 

/.  Mar.  Biol.  Assoc.  U.K.,  21 :  687-704. 
YONGE,  C.  M.,  1946.     On  the  habits  of  Turritclla  comnuinis  Risso.     /.  Mar.  Biol.  Assoc.  U.K., 

26:  377-380. 
ZITTELL,  K.  A.,  1913.     Text  book  of  palaeontology.     Macmillan  and  Co.,  London. 


Reference:  Biol.  Bull..  154:  472-484.     (June,  1978) 


FOOD-RESOURCES  AND  THE  INFLUENCE  OF  SPATIAL  PATTERN 

ON  FEEDING  IN  THE  PHORONID 
PHORONOPSIS  VIRIDIS 

THOMAS  E.  RONAN,  JR. 
Department  of  Earth  and  Space  Sciences,  University  of  California,  Los  Angeles,  California  90024 

The  Phoronida  are  a  coelomate  phylum  of  vermiform,  lophophorate  tube-dwell- 
ing organisms.  Although  the  phylum  consists  of  but  two  genera  and  some  eleven 
species  (Emig,  1974),  all  resident  in  shallow  marine  waters  (Hyman,  1959), 
it  is  possibly  of  great  phylogenetic  and  ecological  importance.  Indeed,  phoronids 
may  well  represent  the  most  primitive  of  living  deuterostomes  (Zimmer,  1964) 
and  the  ancestoral  stock  of  the  lophophorates  (Valentine,  1973;  Farmer,  Valentine 
and  Cowen,  1973).  Ecologically,  phoronids  are  often  important  in  the  structure 
of  soft-sediment  and  fouling  communities  in  that  they  may  monopolize  primary 
space  (Ronan,  1975),  a  potentially  limiting  resource  as  in  the  rocky  intertidal 
region.  Despite  the  significance  of  the  Phoronida,  it  remains  a  relatively  obscure 
phylum  which  has  not  attracted  the  attention  of  many  investigators.  Hyman 
(1959)  has  reviewed  the  literature  pertaining  to  phoronid  biology.  Work  in 
English  on  the  Phoronida  has  emphasized  systematics  (Marsden,  1959;  Emig, 
1974),  developmental  biology  (Rattenbury,  1953;  Zimmer,  1964,  1967),  and 
genetics  (Ayala,  Valentine,  Barr,  and  Zumwalt,  1974)  ;  phoronid  ecology  has 
received  little  attention  (MacGinitie,  1935;  Johnson,  1959;  Ronan,  1975). 

This  paper  examines  the  spatial  pattern,  feeding,  and  food-resources  of  the 
phoronid  Phoronopsis  znridis,  a  large  phoronid  with  a  pale  green  lophophore  which 
inhabits  intertidal  localities  in  west  coast  embayments.  In  the  past  it  has  been 
assigned  to  Phoronopsis  viridis  (Hilton,  1930)  based  on  specimens  from  Morro 
Bay,  California,  but  it  has  also  been  synonymized  with  P.  harmeri  (Pixell,  1912) 
described  from  Vancouver  Island,  British  Columbia  (Marsden,  1959).  Zimmer 
(University  of  Southern  California,  personal  communication)  believes  the  Canadian 
and  Calif ornian  populations  are  specifically  distinct,  with  the  California  form  to 
be  called  P.  viridis;  I  shall  employ  this  name,  although  there  is  no  work  o;n 
geographic  variation  and  relationships  are  uncertain. 

MATERIALS  AND  METHODS 
Study  site 

The  study  was  conducted  in  Bodega  Harbor,  California  (38°  19'  N, 
123°  03'  W),  a  small  marine  coastal  embayment  located  100  kilometers  north  of 
San  Francisco,  California.  The  harbor  is  quite  shallow  (maximum  depth,  4.0  m 
at  low  water).  At  mean  lower  low  water  (MLLW:=0.0  ft)  extensive  tidal 
flats,  which  occupy  about  60%  of  the  harbor,  are  exposed.  The  harbor  is  a 
depositional  environment.  Without  periodic  maintenance  dredging,  the  harbor 
would  revert  to  a  lagoon. 

472 


FEEDING  IN  PHORONOPSIS  473 

Within  the  harbor  there  is  a  0.5  mile2  sand  flat  which  is  posted  and  main- 
tained as  a  marine  life  refuge  by  the  University  of  California  Bodega  Marine 
Laboratory.  Phoronid  and  sediment  collections  were  taken  within  the  refuge  and 
just  to  the  north  of  the  refuge.  Feeding  observations  and  nearest-neighbor  (N-N) 
measurements  were  made  within  the  refuge  boundaries. 

Field  census  and  nearest-neighbor  relations 

The  intertidal  distribution  and  abundance  of  Phoronopsis  were  determined  by 
hand  excavation  of  square  meter  holes  along  two  transects.  The  transects 
were  roughly  parallel  to  each  other  from  MLLW  to  the  mean  higher  high  water 
(MHHW)  mark  (120  cm  above  MLLW).  The  distance  between  transect 
stations  was  10  meters;  the  longer  transect  A  had  twice  as  many  stations  as 
transect  B.  Care  was  taken  to  establish  the  transect  stations  in  areas  known 
to  be  free  from  clam  digging  which  can  greatly  modify  the  spatial  pattern  of 
Phoronopsis.  During  excavation,  all  phoronid  tubes  were  separated  from  the 
sedimentary  matrix  and  their  numbers  recorded.  Phoronid  numbers  were  esti- 
mated at  95 %  of  the  counted  numbers  of  tubes  because  about  5%  of  the  tubes  in 
dense  aggregations  are  known  to  be  vacant  (Ronan,  1975). 

Nearest-neighbor  measurements  were  made  along  transect  A,  at  stations  3, 
5,  and  7,  following  methods  proposed  by  Clark  and  Evans  (1954).  Spacing 
measurements  were  not  possible  at  station  1  because  phoronid  tube  apertures 
were  occluded  by  flocculent  seston  which  thickly  mantled  the  depositional  interface. 
Higher  in  the  intertidal,  the  spatial  pattern  of  Phoronopsis  was  easier  to  determine 
since  the  small  holes  produced  by  the  animal  at  the  sediment-water  interface 
(SWI)  remain  open  at  low  water.  Because  individuals  of  this  species  aggregate 
in  clusters  of  up  to  thousands  per  m2  throughout  the  study  area,  all  N-N  measure- 
ments are  within  cluster  distances.  At  each  sampling  station,  three  25  cm2 
frames  fitted  with  clear  plastic  inserts  were  randomly  dropped  and  the  area 
occupied  by  the  largest  cluster  circumscribed  with  a  rectangle.  Within  the 
rectangle,  the  position  of  each  animal  was  recorded  on  the  plastic  with  a  felt 
tip  marker.  Since  only  inhabited  tubes  had  open  apertures,  cluster  population 
density  was  accurately  determined  by  counting  the  dots  on  the  plastic.  For  all 
animals,  distance  to  N-N  was  estimated  as  the  distance  to  the  nearest  mm  between 
the  centers  of  the  dots. 

Feeding  observations 

Low  intertidal  sites  (2  m2),  each  estimated  to  contain  more  than  17,000 
Phoronopsis,  were  selected  for  detailed  underwater  feeding  observations.  The 
study  sites  were  adjacent  to  transect  A  and  separated  from  each  other  by  5  m. 
About  six  hours  were  spent  underwater  on  various  occasions  observing  phoronid 
feeding  behavior. 

As  a  phoronid  lophophore  is  small  and  held  close  to  the  SWI,  it  is  best  to 
view  it  from  the  side.  Feeding  observations  were  made  by  SCUBA  diving  with  a 
heavy  weight  belt  and  tethering  to  a  short  line  anchored  in  the  sediment  nearby. 
Height  measurements  were  made  on  nine  clustered  phoronids  every  15  minutes 
over  a  60-minute  period. 


474  THOMAS  E.  RONAN,  JR. 

Food  resources 

Early  observations  indicated  that  a  feeding  animal  positions  its  lophophore 
within  the  turbid  near-bottom  layer  of  water  (Ronan,  1975).  The  location  of 
the  feeding  appendage  is  reflected  in  the  animal's  stomach  contents  in  that  the 
ingested  materials  primarily  represent  items  resuspended  from  the  SWI.  To 
confirm  this  impression,  food  selection  in  relationship  to  the  animal's  available 
food  was  quantified  by  examining  the  food-resources  of  the  SWI  and  the  water 
column. 

A  large  diameter  (5  mm)  pipette  was  used  to  collect  seston  (skeletal  material, 
mineral  grains,  and  organic  particles)  from  the  SWI  around  the  tubes.  The 
seston  was  preserved  in  90%  alcohol  and  examined  under  a  dissecting  microscope. 
Using  the  criteria  of  Johnson  (1974),  seston  material  was  classified  by  particle 
type.  Mineral  grains  were  categorized  by  size  and  the  presence  or  absence  of 
encrusting  organic  matter.  Loose  aggregates  of  fine-grained  minerals  bound  in 
an  organic  matrix  were  termed  floe  (organic-mineral)  aggregates.  Firm  organic- 
mineral  aggregates  in  the  form  of  pellets,  or  fragments,  were  classified  as  either 
Phoronopsis  feces  (which  are  distinctive)  or  other  fecal  matter.  The  remaining 
material  was  listed  as  either  plant  fragments,  pollen,  diatoms,  or  small  metazoans 
(copepods,  nematodes,  ostracods,  etc.).  Mineral  grains  were  measured  with  an 
ocular  micrometer.  Particle  type  abundance  categories  (Johnson,  1974)  were 
used  to  express  the  abundance  of  different  fractions  of  the  food-resources  available 
to  the  organism. 

During  the  same  period  (Sept.-Oct,  1975)  in  which  phoronids  were  collected 
for  stomach  content  analysis,  plankton  was  also  collected  by  towing  a  0.25  m 
plankton  net  with  0.333  mm  mesh  size  twice  through  the  water  with  the  base  of 
the  net  no  more  than  10  cm  above  the  bottom  for  two  15-minute  periods.  The 
entire  sample  was  analyzed  and  particle  type  abundance  categories  (which  were 
calculated  by  averaging  the  two  samples)  were  used  to  express  the  relative 
abundance  of  the  plankton  species  available  to  Phoronopsis. 

Stomach  contents 

Twenty  specimens  of  Phoronopsis  were  removed  from  their  tubes,  preserved 
in  90%  alcohol,  and  the  ingested  material  collected  from  the  stomach.  After  the 
stomach  fractions  were  washed  in  distilled  water  to  remove  adhering  mucus,  they 
were  examined  under  magnification.  The  methods  described  above  for  analyzing 
particle  fractions  were  employed  to  determine  abundance  of  particle  types  and 
size  distributions  for  the  stomach  samples. 

The  results  of  the  analysis  of  abundance  of  different  particle  types  are  expressed 
as  percentage  of  particle  abundance.  Whitlatch  (1974)  suggests  the  use  of  this 
measure  in  determining  food  selection  because  it  reflects  the  relative  amounts 
of  different  particles  available  in  the  environment  of  an  organism. 

Electivity  coefficients  of  different  particle  types  selected  by  Phoronopsis  were 
determined  using  the  statistic  of  Ivlev  (1961).  The  statistic  is  calculated  as 
E'  =:  (T!  —  Pi)/(ri  +  PJ).  For  the  ith  food  type,  rt  equals  the  percentage  ingested 
and  pi  is  the  percentage  of  that  food  type  available  in  the  environment.  The 
coefficient  is  bounded  and  symmetrically  distributed  about  zero  (E'  :=  0  indicates 


FEEDING  IN  FIIORONOPSIS 


475 


nonselective  feeding;    -1  ; ;  E'  <  0  indicates  avoidance;  and  0  <  E'    ;  1   indicates 
feeding  preference). 

Phoronid  fecal  pellets  were  collected  from  the  field  with  a  small-diameter 
(2  mm)  pipette,  washed  and  disaggregated  in  sea  water  on  a  250  ju.ni  sieve,  and 
the  contents  examined  under  magnification.  Particle  size  analyses  of  disaggregated 
feces  were  performed  in  distilled  water. 

RESULTS 

Abundance  and  spatial  distribution 

Figure  1  shows  the  number  of  Phoronopsis  excavated  from  meter-square 
quadrats  along  the  two  transects.  In  the  intertidal  zone,  phoronids  are  aggregated 
in  discrete  clusters  that  are  separated  from  other  clusters  by  intervening  open 


25 

IO 

O 

—  20 
X 

C\J 


-£    10 
o 

<      5 


Transect  (A) 


012345678 

Station  No. 


9   10 


Transect  (B) 


25 


20 
X 


^    15 

L. 
o> 

Q. 

%    10 

E 
'c 
<  5 


12345 

Station  No. 


120  r 


Eioo 


MLLW  0 


i      i      i      i 


I     23456789    10 

Station  No. 


120  r 


tlOO 


80 


a> 

>  60 
_a 


o  40 


-S>  20 
LU 


MLLWO 


i      i      i     i 


12345 

Station  No 


FIGURE  1.  A  and  B  (above)  show  the  number  of  Phoronopsis  viridis  removed  from 
square-meter  excavations  along  two  parallel  intertidal  transects ;  and  (below)  station  eleva- 
tions with  respect  to  distance  above  mean  lower  low  water.  The  sampling  interval  between 
stations  was  10  m. 


476 


THOMAS  E.  RONAN,  JR. 


spaces  where  density  is  low.  The  clusters  were  largest  and  the  densities  greatest 
within  firm  sediments  with  a  median  particle  diameter  less  than  250  /xm.  Along 
the  transects  which  traversed  fine  sediment  only,  cluster  density  ranged  from 
21,422  phoronids/nr  at  MLLW  to  zero  at  the  highest  intertidal  stations ;  their 
upper  limit  roughly  corresponded  to  the  mean  lower  high  water  (MLHW  =  :  100 
cm  above  MLLW)  line.  The  mosaic  of  irregular  clusters  and  intervening  open 
spaces  was  most  obvious  near  the  animals'  upper  limit  where  clusters  were  small 
(range:  3-67  phoronids)  and  their  boundaries  distinct.  The  most  dense  and  con- 
tinuous clusters  were  around  MLLW,  where  aggregate  clusters  containing  up  to 
150,000  phoronids  and  covering  up  to  10  m2  occurred. 

Phoronids  were  not  distributed  evenly  throughout  the  low  and  mid-intertidal, 
however.  Phoronopsis  viridis  was  absent  or  rare  in  lenses  of  loosely  packed 
sediment  with  a  median  particle  diameter  greater  than  250  ju.ni.  In  an  elliptical 
bed  of  coarse  sand  (5  X  30  m)  along  transect  A,  located  eight  meters  shoreward 
from  the  low  tide  line,  cluster  density  declined  from  12,116  phoronids/m2  at  the 
periphery  of  the  bed  to  0/m2,  1.5  meters  into  the  interior  of  the  coarse  substrate. 
A  smaller  lens-shaped  (5  X  8  m)  body  of  coarse  sediment  occurred  in  the  mid- 
intertidal.  In  this  substrate,  the  cluster  density  of  phoronids  was  very  low,  ranging 
from  9370  animals/m2  at  the  periphery  to  0  animals/m2  at  a  distance  of  0.5  m 
into  the  interior.  Although  the  number  of  animals  per  square  meter  always 
declined  in  loose  coarse  sediment,  the  phoronids  remain  aggregated  in  small,  tight 
clusters. 

Table  I  shows  the  N-N  frequency  distribution  for  2616  Phoronopsis  in  nine 
intertidal  clusters  composed  of  from  75  to  481  individuals.  All  of  the  measure- 
ments were  made  in  fine  sediment  prior  to  the  excavation  of  stations  3,  5,  and  7 
along  transect  A.  The  mean  distance  to  N-N  was  5.4  ±  2.8  (s.d.)  mm  with 
a  range  of  1-25  mm.  The  mode  is  slightly  less  than  the  mean,  and  the  dis- 
tribution has  a  long  tail  of  distances  greater  than  the  mean. 

N-N  analysis  of  the  dispersion  pattern  of  Phoronopsis  indicates  the  distribution 


TABLE  I 

Frequency  distribution  of  distances  between  individuals  for  2616  Phoronopsis  viridis  in  nine  intertidal 
aggregations. 


Distance  (to  nearest  mm) 

Frequency 

Per  cent 

Cumulative  percentage 

1 

92 

3.52 

— 

2 

183 

7.00 

10.52 

3 

401 

15.33 

25.85 

4 

516 

19.72 

45.57 

5 

360 

13.76 

59.33 

6 

341 

13.04 

72.37 

7 

293 

11.20 

83.57 

8 

125 

4.78 

88.35 

9 

136 

5.20 

93.55 

10 

33 

1.26 

94.81 

11-13 

79 

3.01 

97.82 

14-16 

36 

1.37 

99.19 

>17 

21 

0.80 

99.99 

FEEDING  IN  PHORONOPSIS  477 

of  individuals  within  clusters  departed  significantly  from  randomness.  A  mean 
R  (R)  value  for  9  clusters  (cluster  size  75-481  individuals;  N  ==  2616)  of  0.64 
indicates  that  N-N  were  only  slightly  more  than  half  as  far  apart  as  expected  under 
conditions  of  randomness  (P  <  0.01).  The  mean  N-N  distance  of  5.4  mm  was 
roughly  half  the  space  required  to  expand  two  adjacent  lophophores  completely, 
without  their  impinging  each  other.  Individuals  crowded  together  in  this  manner 
were  observed  to  stratify  their  lophophores  vertically  by  some  individuals  extend- 
ing their  trunk  above  the  SWI.  Because  the  body  wall  is  flexible,  a  phoronid  can 
spread  the  lophophore  away  from  neighbors  by  bending  the  extended  trunk.  Small 
N-N  distances  often  produced  an  array  of  "tall"  and  "short-standing"  phoronids 
as  the  individuals  in  a  cluster  maneuver  for  space  to  expand  their  lophophores. 
However,  few  (<  11%)  individuals  of  Phoronopsis  were  packed  within  3  mm  of 
each  other.  At  such  close  distances,  even  a  stratification  of  feeding  appendages 
failed  to  provide  for  full  expansion  of  adjacent  lophophores;  the  N-N  distances 
were  so  slight  that  an  expanded  lophophore  would  abut  against  the  trunk  of  a 
neighboring  "tall-standing"  individual.  Therefore,  the  distance  from  the  lophophore 
of  one  individual  to  the  trunk  of  an  adjacent  individual  provides  a  measure  of  the 
lower  limit  below  which  feeding  space  cannot  be  reduced  by  a  stratification  of 
lophophores. 

The  lophophore 

The  action  of  the  lophophore  was  observed  microscopically  in  laboratory- 
maintained  Phoronopsis.  Currents  produced  by  cilia  on  the  tentacles  bring 
water  and  suspended  particles  down  within  the  loop  of  the  lophophore  and  then 
out  between  the  tentacles.  The  mechanics  of  particle  capture,  rejection,  and  food 
transport  were  found  to  be  as  described  for  Phoronis  vancouverensis  by  Strath- 
mann  (1973). 

There  was  no  diel  periodicity  in  feeding ;  animals  fed  continuously  during 
tidal  submergence.  This  observation  was  expected,  since  variations  in  food  avail- 
ability are  not  to  be  expected.  Although  the  lophophore  is  perpetually  bombarded 
by  small  particles  cascading  along  the  SWI,  the  animal  is  sensitive  to  dis- 
turbance and  has  a  well-developed  escape  response.  Contact  between  the  tentacles 
of  the  lophophore  and  large  tumbling  fecal  pellets  and  detritus  produced  a  partial 
folding  of  the  lophophore  and  retraction  of  the  trunk.  Predatory  strikes  at  the 
feeding  appendage  by  the  nudibranch  Hermissenda  crassicornis  resulted  in  a 
rapid  folding  of  the  lophophore  and  retraction  into  the  tube.  Only  rarely  was 
a  large  lophophore  completely  cropped  by  Hermissenda;  most  strikes  removed 
only  a  few  tentacles  from  the  lopophore.  However,  a  small  phoronid  can  lose 
its  entire  lophophore  to  Hermissenda  as  well  as  to  fishes  which  can  remove  even 
a  large  phoronid  from  its  tube  (Ronan,  in  preparation). 

Feeding  in  clusters 

Clustered  phoronids  were  observed  to  sporadically  vary  the  height  of  their 
expanded  lophophores.  The  structure  of  the  stratification  undergoes  continuous 
modification  as  the  animals  raise,  lower,  and  interfinger  their  feeding  appendages 
(Table  II). 


478 


THOMAS  E.  RONAN,  JR. 


TABLIC  II 

Sequential  underwater  measurements  of  the  height  of  Phoronopsis  viridis  above  Hie  sediment-water 
interface. 


Time  (min) 

0 

15 

30 

45 

60 

X 

1 

15 

0 

8 

13 

17 

11 

2 

24 

24 

20 

22 

25 

23 

3 

18 

23 

12 

0 

9 

12 

4 

19 

21 

6 

0 

14 

12 

5 

14 

8 

8 

5 

9 

9 

6 

22 

22 

16 

8 

3 

14 

7 

s\  9 

0 

11 

14 

9 

9 

8 

F17 

17 

24 

19 

19 

19 

9 

'  24 

20 

25 

16 

0 

17 

Mean  height*  of  cluster 

18 

15 

14 

11 

12 

*  Height  measurements  were  made  by  inserting  a  transparent  metric  ruler  directly  in  front  ot 
the  animal  and  recording  the  distance  from  the  sediment-water  interface  to  the  base  of  the  lopho- 
phore.  The  mean  distance  to  nearest-neighbor  was  4.9  mm  with  a  range  of  2-13  mm. 

The  longest  trunk  extension  noted  was  25  nun  with  the  lophophore  extending 
another  6  mm  above  the  trunk.  The  trunk  is  flexible  so  than  an  animal  may  bend 
away  from  and  expand  the  feeding  appendage  above  surface  obstructions.  Periodic 
height  adjustments  maintain  the  stratification  and  safeguard  against  impingement 
between  neighboring  lophophores. 


Scston  composition 

Analysis  of  the  food-resources  of  the  SWI  revealed  that  three  particle  types 
averaged  over  69%  of  the  potential  available  food:  small  (<  100  /xm)  encrusted 
mineral  grains,  floe  aggregates,  and  Phoronopsis  fecal  pellets  (Table  III).  Small 
encrusted  mineral  grains  were  usually  the  most  abundant  particle  type.  The 
encrusting  material  varied  in  its  consistency  and  degree  of  adherence.  When 
the  organic  matter  attached  to  mineral  grains  is  stained  with  the  periodic  acid- 
Schiff  (PAS)  histological  reagent,  it  characteristically  gives  a  strong  positive 
reaction,  thus  suggesting  the  encrusting  material  is  largely  carbohydrate  (John- 
son, 1974;  Whitlatch,  1974).  Small  encrusted  mineral  grains  plus  floe  aggregates, 
the  second  most  abundant  particle  type,  together  averaged  more  than  56%  of 
the  available  participate  material. 

Floe  (organic-mineral)  aggregates  comprised  the  second  most  abundant  par- 
ticle type  in  the  samples  examined.  Floe  material  consists  of  very  fine-grained 
mineral  matter,  incorporated  into  an  amorphous  organic  matrix  (Johnson,  1974; 
Whitlatch,  1974).  Not  all  floe  material  is  the  same.  Some  aggregates  were  rich 
in  mineral  matter,  tightly  bound  by  the  matrix  material.  Other  floe  material 
consisted  of  a  loose  indistinct  matrix  with  few  bound  participates.  From  exten- 
sive staining  experience,  Johnson  (1974)  and  Whitlatch  (1974)  conclude  that 


FEEDING  IN  PHORONOPSIS  479 

the  matrix  of  organic-mineral  aggregates  is  largely  carbohydrate.  Floe  material 
was  always  abundant  in  harbor  water  samples  collected  from  just  above  the  SWI 
and  was  especially  abundant  in  samples  collected  from  the  low  intertidal  zone  and 
tidal  channels.  Rhoads  (1973)  has  reported  that  different  types  of  floe  material 
may  differ  in  Hoc  bulk  density  and  ease  of  resuspension. 

Fragmented  fecal  pellets  of  Phoronopsis  were  the  third  major  particle  type. 
Intact  pellets  are  spindle-shaped  rods  up  to  7  mm  in  length,  which  are  rich  in  silt 
and  clay.  Natural  decomposition  of  the  mucous  envelope  which  binds  a  pellet 
produces  many  stringy  fecal  fragments.  There  was  a  strong  morphological  re- 
semblance between  naturally  decomposing  phoronid  fecal  matter  and  the  floe 
material  complexed  with  mineral  grains  (organic-mineral  aggregates). 

All  three  common  categories  of  seston  (small  encrusted  mineral  grains,  floe 
aggregates,  and  fecal  pellets)  were  resuspended  by  tidal  currents  and  wind-driven 
waves.  Hence  they  were  readily  available  to  Phoronopsis.  The  remaining  par- 
ticulate  material  consists  of  large  mineral  grains  (>  100  /xm),  plant  detritus 
(fragments  of  Ulva  c.vpansa  and  Zostcra  marina),  living  diatoms,  pollen,  and  a 
variety  of  small  metazoans  (copepods,  ostracods,  nematodes,  etc.). 

Plankton  composition 

During  fall  sampling,  there  was  a  plankton  bloom  in  the  harbor.  A  pair  of 
daytime  plankton  tows  from  about  10  cm  above  the  phoronid  bed  contained 
approximately  35%  dinoflagellates  (Ccratium  sp.  and  Gonyaulax  sp.),  12%  centric 
diatoms  (two  species  each  of  Chactoceros  sp.  and  Coscinodiscns  sp.),  10% 
harpacticoid  copepods,  7%  Cancer  crab  zoea,  6%  ostracods  (?  Cylindroberis  sp.), 
and  4%  hydromedusae  (Polyorchis  sp.).  The  remaining  living  material  con- 
sisted mostly  of  pennate  diatoms  (1.9%)  and  small  flagellates  (1.3%). 

The  tow  also  contained  two  types  of  organic  detritus  that  constituted  about 
18.8%  of  the  samples;  amorphous  strings  and  balls  of  organic  matter  and 
Zostera  marina  fragments  averaged  11  and  7%  of  the  samples,  respectively. 

TABLE  III 
Particle  type  abundance  of  seston*  sampled  near  the  tubes  of  Phoronopsis  viridis,  as  mean  percentages . 

Particle  type  x 

Mineral  100-200  Mm  encrusted  7.0 

Mineral  100-200  /urn  not  encrusted  6.6 

Mineral  <  100  /urn  encrusted  35.0 

Floe  aggregates  21.1 

Plant  detritus  3.1 

Pollen  grains  1.0 

P.  viridis  fecal  fragments  13.4 

Other  fecal  matter  7.1 
Living  diatoms 

Small  metazoans  2.1 

*  Seston  is  denned  as  inorganic  detritus  and  organic  (living  and  nonliving)  particles.  Mean 
percent  abundance  was  determined  by  counting  and  averaging  200  particles  at  each  of  eight 
sampling  stations. 


480  THOMAS  E.  RONAN,  JK. 

TAULK  IV 

Electivity  coefficients  of  seven   most  (i/nintlunt  particle  types  in  the  .stomach  of  Phoronopsis  viridis 
(data  averaged  for  20  animals). 

Particle  type  Electivity 

Mineral  100-200  Mm  (encrusted)  -0.18 

Mineral  100-200  ^m  (not  encrusted)  —0.15 

Mineral  <  100pm  (encrusted)  +0.11 

Floe  aggregates  +0.08 

Dinoflagellates  +0.05 

P.  viridis  feces  +0.03 

Diatoms  -0.20 


Stomach   contents 

Only  six  items  were  routinely  present  in  the  stomach  of  Phoronopsis.  Positive 
electivity  values  suggest  a  preference  for  small  (<  100  /mi)  encrusted  mineral  grains 
(Table  IV).  Within  this  category,  35-75  /mi  mineral  grains  were  thickly  en- 
crusted with  loosely  adhering  organic  matter.  The  electivity  data  also  indicate 
a  preference  for  floe  aggregates  along  with  planktonic  dinoflagellates.  In  60%  of 
the  animals  examined,  floe  aggregates  of  silt-  and  clay-sized  materials  occupied 
over  one-third  the  volume  of  the  stomach.  Small  dinoflagellates  were  selected 
most  often. 

A  strong  avoidance  was  displayed  for  mineral  grains  larger  than  100  /mi. 
Organic  encrustations,  which  increase  both  the  sphericity  and  effective  diameter  of 
the  particles,  further  reduced  the  electivity  of  large  mineral  grains.  Avoidance  of 
particles  in  the  100-220  jwm  range  probably  is  due  to  either  the  inability  of  the 
frontal  cilia  on  the  tentacles  to  transport  the  particles  or  an  upper  limit  to  the 
size  of  material  which  can  be  ingested. 

Fecal  pellets 

The  seston  and  plankton  ingested  by  the  animal  is  defecated  at  the  SWI  as 
easily  fragmented  fecal  pellets.  Embedded  in  the  fine-grained  mucous  matrix 
were  mineral  grains  (50—90  ^m)  and  an  occasional  pollen  grain  or  still  motile 
ciliate.  The  common  mulibranch,  Hennisscnda  crassicornis,  was  observed  to 
ingest  large  numbers  of  phoronid  fecal  pellets.  The  importance  of  fecal  material 
as  a  food  source  for  invertebrates  has  been  demonstrated  by  Newell  (1965)  and 
Johannes  and  Satomi  (1966).  They  have  shown  that  the  bacteria  which  decompose 
feces  are  more  important  nutritionally  than  the  waste  material  present. 

Some  of  the  phoronid  fecal  material  is  incorporated  into  the  sediment  by 
numerous  small  burrowing  metazoans  which  disaggregate  and  intermix  fecal  ma- 
terial with  the  surface  sediment.  Floe,  or  organic-mineral  aggregates,  is  probably 
produced  mainly  by  the  mixing  of  decomposing  phoronid  (or  other)  fecal  material 
and  sediment.  Unmixed  fecal  material  accumulates  in  surface  depressions  (ripple 
troughs,  ray  feeding  pits,  etc.),  decomposes,  and  becomes  flocculent  seston. 
Resuspension  of  this  material  makes  it  available  for  ingestion  by  Phoronopsis. 


FEEDING  IN  PPIORONOPSIS  481 

DISCUSSION 

In  recent  years  a  number  of  studies  have  been  made  of  the  distribution 
patterns  of  benthic  species.  Most  studies  suggest  that  distributions  tend  toward 
aggregation  and  that  random  or  uniform  distributions  seldom  occur  in  marine 
(e.g.,  Clark  and  Milne,  1955;  Angel  and  Angel,  1967;  Warner,  1971)  or 
terrestrial  environments  (Greig- Smith,  1964;  Pielou,  1969).  Surprisingly, 
although  a  number  of  soft-sediment  species  are  known  to  form  dense  aggregations, 
particularly  brittle  stars  (Warner,  1971  ;  Broom,  1975 ;  Wilson,  Holme,  and 
Barrett,  1977),  there  is  little  statistical  information  on  the  distribution  of  indi- 
viduals within  such  aggregations. 

The  present  study  provides  detailed  statistical  information  on  Phoronopsis, 
which  forms  dense  aggregations  in  the  intertidal  region.  Detailed  sampling  has 
shown  that  the  population  exhibits  a  clumped  distribution  whose  degree  of 
aggregation  remains  relatively  constant  with  changes  in  intertidal  elevation  and 
population  density.  This  close  association  between  nearest  neighbors  produces  a 
pattern  of  tight  clusters. 

These  results  differ  from  those  reported  by  Johnson  (1959)  who  has  used 
another  N-N  measure  (Clark  and  Evans,  1955)  to  examine  the  spatial  pattern  of 
Phoronopsis.  His  results  indicate  that  individual  animals  tend  to  be  distributed 
evenly  within  clusters.  Further,  he  suggests  that  this  pattern  of  dispersion 
reflects  the  minimum  distance  between  individuals  necessary  for  feeding,  but  he 
reports  no  N-N  distances  nor  does  he  mention  a  stratification  of  feeding  appendages. 

In  general,  invertebrates  that  commonly  form  large,  dense  aggregations  are 
animals  that  spend  much  of  their  time  suspension  feeding  (e.g.,  Ophiothrix 
jragilis,  Warner,  1971  ;  Dcndrastcr  e.vcentris,  Timko,  1975  ;  Spisula  sol  id  a,  Ford, 
1925;  Ampelisca  spp.,  Mills,  1967).  This  emphasis  on  feeding  activity  means 
that  they  are  continually  placed  in  situations  that  expose  them  to  disturbance  and 
probably  make  them  highly  susceptible  to  predators.  Although  a  close  association 
between  phoronids  creates  spacing  problems  among  themselves  for  expansion  of 
the  lophophore  during  feeding,  clustering  may  be  an  adaptation  to  predation :  when 
N-N  distances  are  small  and  lophophores  stratified,  a  close  association  between 
individuals  can  limit  the  number  of  animals  available  to  the  predatory  nudibranch 
Hcrmissenda  crassicornis.  I  have  observed  that  the  sudden  retraction  of  a 
lophophore  creates  a  disturbance  that  is  transmitted  to  neighboring  animals  either 
by  collision  of  overlapping  lophophores  or  by  the  generation  of  sudden  perceptible 
pressure  waves  that  can  produce  multiple  retraction  of  lophophores.  Although  the 
clusters  are  noncolonial  aggregations,  this  imperfect  wave  of  withdrawal  that  spreads 
over  part  of  the  cluster  produces  a  response  that  makes  the  cluster  less  vulnerable 
to  predation.  Without  the  response,  escape  of  Phoronopsis  would  depend  upon 
contact  with  a  crawling  predator  such  as  Hcrmissenda,  which  could  more  easily 
forage  through  the  cluster. 

A  dense  assemblage  of  Phoronopsis  can  also  stabilize  sediment  and  limit  bur- 
rowing of  large  errant  infauna  which  are  potentially  destructive  to  the  phoronids. 
In  areas  of  natural  contact  between  the  thallassinid  sandshrimp  Caltianassa  cali- 
fornicnsis  and  Phoronopsis,  the  burrowing  activity  of  the  shrimp  can  act  to  set 
the  upper  limit  of  Phoronopsis  intertidal  range  (Ronan,  1975).  This  type  of 


482  THOMAS  E.  RONAN,  JR. 

interaction  in  which  one  population  is  limited  while  the  other  is  not  has  been 
termed  "amensalism"  (Odum,  1971).  While  there  is  no  evidence  of  shrimp  pre- 
clation  at  low  population  densities  of  Phoronopsis,  manipulated  tuhes  are  fre- 
quently found  at  unnatural  depths,  and  occasionally  tuhes  are  found  to  he  actually 
hroken  with  pieces  of  tube  offset  and/or  rotated  on  opposite  sides  of  Callianassa 
burrows.  Former  occupants  of  broken  and  disoriented  tubes  were  found  to  be 
living  free  in  the  sediment  in  the  process  of  building  new  tubes  in  contact  with 
the  water  column.  This  nonpredatory  but  potentially  destructive  interaction  with 
Callianassa  constitutes  a  form  of  "substrate  amensalism"  that  operates  at  low 
tube  densities  to  restrict  the  intertidal  distribution  of  Phoronopsis.  However, 
when  Phoronopsis  densities  are  high  and  N-N  distances  small,  the  numerous 
tubes  buttress  the  sediment  and  constitute  a  subsurface  obstruction  to  some  large 
burrowing  organisms.  Dense  clusters  of  Phoronopsis  are  only  rarely  undermined 
by  foraging  Callianassa  (Ronan,  in  preparation). 

Cluster  formation  may,  therefore,  permit  Phoronopsis  to  coexist  in  sandilats 
with  an  established  errant  infauna  which  it  might  not  otherwise  successfully  inhabit. 
However,  cluster  formation  also  could  have  other  advantages :  first,  the  proximity 
of  large  numbers  of  adults  could  insure  gamete  fertilization  during  the  breeding 
season ;  and  secondly,  clustering  may  have  even  more  subtle,  advantageous  effects 
on  feeding.  The  feeding  currents  of  an  individual  may  work  better  with  other 
individuals  nearby.  Aggregated  feeding  currents  may  possibly  modify  localized 
water  flow  with  the  clusters  acting  as  "food  funnels"  for  the  accumulation  of 
both  resuspended  and  planktonic  food  material.  The  thick  seston  layer  which 
develops  within  phoronid  clusters,  but  not  in  the  open  spaces  between  clusters, 
may  be  a  manifestation  of  the  funneling  effect. 

Previous  reports  of  diet  composition  and  selectivity  in  the  Phoronida  are 
lacking.  However,  there  are  studies  which  are  pertinent  to  the  present  work. 
Whitlatch  (1974  and  personal  communication)  has  shown  that  the  polychaete 
Pectinaria  yonldii  concentrates  organic  material  found  in  the  sediment  by  pre- 
ferentially ingesting  large  encrusted  mineral  grains,  fecal  material,  and  floe 
aggregates.  He  suggests  that  there  are  probably  several  major  sources  of  the 
organic  material  that  encrusts  mineral  grains  and  forms  low-density  floe  aggregates 
(terrigenous  input,  plant  debris,  decomposing  fecal  material,  and  metabolites  of 
plankton  and  bacteria)  and  food  value  differences  may  depend  upon  the  original 
source,  state  of  decay,  and  number  of  times  the  material  has  passed  through  an 
animal  gut.  Further,  he  has  demonstrated  that  the  feeding  of  Pectinaria  channels 
large  amounts  of  organic  material  to  the  SWI  where  it  can  become  available  to 
other  organisms.  At  the  depositional  interface,  the  combined  effects  of  bioturba- 
tion  and  tidal  energy  create  a  constant  upwelling  and  recycling  of  organic  material 
from  the  sediment  into  the  water  column  (Rhoads,  1973).  The  data  on  size 
selectivity  and  diet  presented  in  the  present  report  show  that  resuspended 
encrusted  mineral  grains,  fecal  pellets  and  floe  materials,  and  plankton  are 
of  trophic  significance  to  a  suspension-feeding  phoronid.  The  continuous  feeding 
and  stable  generalized  diet  are  undoubtedly  important  factors  which  have  allowed 
Phoronopsis  to  attain  great  abundance  in  shallow  water  coastal  embayments. 


FEEDING  IN  PHORONOPSIS  483 

I  am  most  grateful  to  Dr.  James  W.  Valentine  for  an  introduction  to  the 
Phoronida  and  helpful  discussion  and  advice.  It  is  a  pleasure  to  acknowledge 
Drs.  R.  I).  \Yhitlatch,  J.  G.  Morin,  R.  R.  Vance,  J.  Standing,  and  an  anonymous 
reviewer  for  reading  and  improving  the  manuscript.  I  wish  to  thank  also 
Dr.  C.  Hand,  J.  Tinkess,  and  the  staff  of  the  University  of  California  Bodega 
Marine  Laboratory  for  providing  space  and  facilities.  This  research  was  partly 
supported  by  ERDA-BLM  contract  E  (4G-3J-34  to  I.  R.  Kaplan  and  W.  E. 
Reed,  University  of  California,  Los  Angeles. 

SUMMARY 

1.  In  the  intertidal  zone  of  Bodega  Harbor,  California,  the  phoronid,  Phoronop- 
sis  rindis,  aggregates  in  clusters  often  composed  of  thousands  of  tightly  aggregated 
individuals    (up    to    150,000/m2).      Within    a   dense   cluster,    there   is    a   spacing 
problem  for  expansion  of  the  lophophores.     When  nearest-neighbor  distances  are 
small,  a  stratification  of  feeding  appendages  is  a  workable  solution  to  the  spacing 
problem,   allowing  simultaneous  expansion   of  clustered   feeding   appendages. 

2.  Suspension-feeding  specimens  of  Phoronopsis  expand  their  lophophores  and 
collect  food  items  from  the  turbid  near-bottom  layers  of  \vater.     Comparison  of 
ingested  items  with  material  collected  where  the  phoronids  feed  indicates  a  prefer- 
ence for  small  (<  100  ^m)  organic  encrusted  mineral  grains,  floe  aggregates,  and 
fecal  material,  all  resuspended  from  the  depositional  interface.     Also  taken  to  a 
lesser  extent  are  plankton  bloom  species,  such  as  diatoms  and  dinoflagellates. 

3.  The  fact  that  Phoronopsis  forms  dense  assemblages  in  the  intertidal  zone 
has  consequences  when  the  community  structure  of  sandtlat  areas  is  considered. 
Although  it  is  probable  that  no  single  factor  can  explain  aggregation  in  Phoronop- 
sis, two  possible  factors,  constituting  strong  selection  pressures  for  cluster  forma- 
tion,   are   relative   immunity   from   disturbance   by   large   burrowing   infauna   and 
protection  from  predation  by  crawling  predators. 

LITERATURE  CITED 

ANGEL,   H.   H.,   AXD  M.   V.   ANGEL,    1967.     Distribution   pattern   analysis   in   a   marine  benthic 

community.     Hcl<jol.  ll'iss.  Mccrcsitntcrs.,  15:  445-454. 
AYALA,  F.  J.,  J.  W.  VALENTINE,  L.  G.  BARR,  AND  G.  S.  ZUMWALT,  1974.     Genetic  variability 

in  a  temperate  intertidal  phoronid,  Phoronopsis  I'iridis.     Biochcin.  Genet.,  11:  413-427. 
BROOM,  D.  M.,  1975.     Aggregation  behaviour  of  the  brittle  star,  Ophiothrix  jniyilis.     J.  Mar. 

Dial.  Assoc.  U.K.,  55  :  191-197. 
CLARK,  P.  J.,  AND  F.  C.   EVANS,   1954.     Distance  to  nearest-neighbor  as  a  measure  of  spatial 

relationships  in  populations.    Ecology,  35  :  445—453. 

CLARK,  P.  J.,  AND  F.  C.  EVANS,   1955.     Some  aspects  of  spatial   pattern   in   biological   popula- 
tions.   Science,  121  :  397-398. 
CLARK,  R.  B.,  AND  A.   MILXE,   1955.     The  sublittoral  fauna  of  two  sandy  bays  on  the  isle  of 

Cumbrae.  Firth  of  Clyde.    /.  Mar.  Biol.  Assoc.  U.K.,  34:  161-180. 
EMIG,  C.  C.,   1974.     The  systematics  and  evolution  of  the  phylum   Phoronida.     Z.  Zool.  Syst. 

Evolutlunsforsch.,  12:  128-151. 
FARMER,  J.   D.,  J.  W.  VALENTINE,  AND  R.   CONVEX,   1973.     Adaptive  strategies  loading   to  the 

ectoproct  ground-plan.    Syst.  Zool.,  22  :  233-239. 
FORD,   E.,    1925.      On   the   growth   of   some   lamellibranchs   in    relation   to   the    food    supply   of 

fishes.     /.  Mar.  Biol.  Assoc.  U.K.,  13:  531-561. 
GREIG-SMITH,  P.,  1964.     Oiiautitatii'c  plant  ccoloi/y,  2nd  cd.     Butterworths,  Washington,   D.C., 

198  pp. 


484  THOMAS  E.  ROMAN,  JR. 

HVMAN,   L.    H.,    1959.      The    invertebrates;    smaller   coelomatc    i/raitps.     McGraw-Hill,    New 

York,  783  ]>p. 
IVI.KV,   V.   S.,    1961.     Ji.rperimeiitdl   ecology   of   the   f  ceiling   of   fixlies.     Yale    University    Press, 

New  Haven,  Connecticut,  302  pp. 
JOHANNES,  R.  E.,  AND  M.  SATOMI,  1966.     Composition  and  nutritive  value  of  fecal  pellets  of  a 

marine  crustacean.     Limnol.   Oceanogr.,   11  :    191-197. 

JOHNSON,  R.  G.,  1959.     Spatial  distribution  of  Phoronopsis  viridis  Hilton.     Science,  129  :   1221. 
JOHNSON,  R.  G.,  1974.     Particulate  matter  at  the  sediment-water  interface  in  coastal  environ- 
ments.   /.  Mar.  Res..  32  :  313-330. 
MAC&NITIE,  G.  E.,  1935.     Ecological  aspects  of  a  California  marine  estuary.     Am.  Mid.  Nat., 

16:  629-675. 
MARSDEN,  J.  R.,  1959.     Phoronidea  from  the  Pacific  Coast  of  North  America.     Can.  J.  Zool., 

37:  87-111. 
MILLS,   E.  L.,   1967.     The  biology  of  an  ampeliscid  amphipod  crustacean   sibling  species  pair. 

/.  Fish.  Res.  Board  Can.,  24  :  305-355. 
NEWELL,  R.  C.,  1965.     The  role  of  detritus  in  the  nutrition  of  two  deposit  feeders,  the  proso- 

branch  H \drohia  uli'ae  and  the  bivalve  Maconin  halthica.     Proc.  Zool.  Soc.  London, 

144:  25-45. 

ODUM,  E.  P.,  1971.    Fundamentals  of  ecology,  3rd  ed.     Saunders,  Philadelphia,  514  pp. 
PIELOU,   E.    C.,    1969.     An    introduction    to   mathematical   ecologv.      John    Wiley,    New    York, 

286  pp. 
RATTENBURY,   J.   C.,    1953.      Reproduction   in   Phoronopsis   viridis.      The    annual    cycle    in   the 

gonads;   maturation  and  fertilization  of  the  ovum.     Biol.   Bull.,    104:    182-196. 
RHOADS,  D.  C.,  1973.     The  influence  of  deposit  feeding  benthos  on  water  turbidity  and  nutrient 

recycling.    Am.  J.  Sci.,  273  :  1-22. 

RONAN,  T.  E.,  JR.,  1975.     Structural  and  paleoecological  aspects  of  a  modern  soft-sediment  com- 
munity :   an  experimental   field   study.     Ph.D.   Dissertation,    University    of   California, 

Davis,  220  pp. 
STRATH  MANN,    R.,    1973.      Function    of    lateral    cilia    in    suspension    feeding    of    lophophorates 

(Brachiopoda,  Phoronida,  Ectoprocta).    Mar.  Biol..  23:  129-136. 
TIMKO,   P.    L.,    1975.      Sand   dollars   as   suspension   feeders:   a   new   description   of   feeding    in 

Dcndrastcr  cxccntris.    Biol.  Bull.,  151  :  247-259. 

VALENTINE,  J.  W.,  1973.     Coelomate  superphyla.     Syst.  Zool.,  22:  97-102. 
WARNER,  G.  F.,   1971.     On  the  ecology  of  a  dense  bed  of  the  brittle  star  Ophiotlirix  fragilis. 

J.  Mar.  Biol.  Assoc.  U.K.,  51  :  267-282. 
WHITLATCH,   R.   B.,   1974.     Food-resource   partitioning  in  the  deposit-feeding  polychaete   Pec- 

tinaria  youldii.    Biol.  Bull.,  147  :  227-235. 
WILSON,  J.  B.,  N.  A.  HOLME,  AND  R.  L.  BARRETT,   1977.     Population  dispersal   in  the   brittle 

star  Opliioconnna  nigra.     J.  Mar.  Biol.  Assoc.   U.K.,  57  :  405-439. 
ZIMMER,  R.  L.,  1964.    Reproductive  biology  and  development  in  Phoronida.    Ph.D.  Dissertation, 

University  of   Washington,   Seattle,  415   pp. 
ZIMMER,  R.  L.,   1967.     The  morphology  and  function  of  accessory  reproductive  glands  in  the 

lophophore  of  Phoronis  Vancouver  ensis  and  Phoronopsis  hanncri.     J .  Morphol.,  121  : 

159-178. 


Reference:  Diol.  Bull.,  154:  485-496.     (June,  1978) 


THE  LIFE  CYCLE  OF  CORYMORPHA   (  =  z  EUPHYSORA)  BIGELOWI 

(MAAS,  1905)  AND  ITS  SIGNIFICANCE  IN  THE  SYSTEMATICS 

OF  CORYMORPHID  HYDROMEDUSAE 

CLAY  SASSAMAN  *  AND  JOHN  T.  REES  - 

Department  of  Biological  Sciences,  Stanford  University,  Stanford,  California  94305; 
and  Bodeya  Marine  Laboratory,  University  of  California, 
Bodega  Bay,  California  94923 

The  systematic  interrelationships  of  medusae  and  polyps  in  the  hydrozoan 
family  Corymorphidae  are,  as  yet,  unclear  in  their  details.  The  metagenic  nature 
of  the  life  cycle  of  Corymorpha  nutans  Sars  was  among  the  first  such  cycles 
described  (Sars,  1835),  but  progress  toward  a  unified  systematics  of  the  family 
has  been  slow.  In  its  original  usage,  the  polyp  genus  Corymorpha  included  a 
heterogeneous  mixture  of  several  separate  evolutionary  lines.  Kramp  (1949) 
proposed  that  polyps  previously  collected  under  this  name  were  of  at  least  two 
lineages,  one  associated  with  the  medusa  genus  Euphysa  Forbes  (1848)  and  the 
other  associated  with  the  medusa  genus  Steenstrupia  Forbes  (1846).  He  resur- 
rected the  polyp  genus  Heteractis  (Allman,  1872)  for  polyps  which  differ  from 
typical  Corymorpha  by  their  permanently  capitate  oral  tentacles,  their  strongly 
contractile  moniliform  aboral  tentacles,  their  lack  of  a  parenchymous  diaphragm 
separating  the  hypostome  from  the  hydrocaulus,  their  replacement  of  a  basal  tuft 
of  root-filaments  by  a  belt  of  papillae  in  the  upper  hydrocaulus,  and  by  differences 
in  perisarc  structure  (Kramp,  1949,  p.  185).  A  compelling  argument  for  the 
subdivision  of  the  family  was  the  observation  that  the  medusae  produced  by 
Heteractis  polyps  were  invariably  species  of  Eitphysa  and  that  the  only  medusa 
known  from  typical  Corymorpha  polyps  was  a  Steenstrupia.  However,  since 
all  Euphysa  medusae  had  not  been  linked  to  Heteractis  polyps  and  all  metagenic 
Corymorpha  polyps  had  not  been  associated  directly  with  Steenstrupia  medusae, 
Kramp  retained  a  dual  classification  system  under  which  the  specific  name  was 
shared  by  both  life  cycle  stages,  but  the  polyp  and  medusae  retained  their 
"classical"  generic  designations.  This  taxonomic  device  was  accepted  by  some 
systematists  (e.g.,  Russell,  1953),  but  has  been  disputed  by  others  (Rees,  1957; 
Naumov,  1960;  Brinckmann-Voss,  1970).  Despite  differences  of  opinion  on 
matters  of  nomenclature,  the  separation  of  the  family  Corymorphidae  into  distinct 
lines  is  generally  accepted.  Indeed,  Rees  (1957)  recognized  four  sub-families 
(Euphysinae,  Corymorphinae,  Boreohydrinae,  and  Branchiocerianthinae),  thereby 
re-establishing  the  Euphysinae  of  Haeckel  (1879)  to  emphasize  the  distinctiveness 
of  the  Euphysa  (--Heteractis}  line. 

Within  this  systematic  framework,  the  position  of  the  genus  Enpliysora  has 

1  Present  address :  Department  of  Biology,  University  of  California,  Riverside,   California 
92521. 

2  Present   address :   Energy   Environment    Division,    Lawrence   Berkeley   Laboratory,    Uni- 
versity of  California,  Berkeley,  California  94720. 

485 


486  C.  SASSAMAN  AND  J.  T.  REES 

remained  totally  obscure.  Since  its  erection  by  Maas  in  1905,  tin's  medusa  genus 
has  had  a  complex  taxonomic  history.  It  has  at  various  times  been  combined 
with  either  Enphysa.  Corymorpha  (as  Stccnstntpia ) ,  or  both.  Hartlaub  (1907) 
immediately  reassigned  the  type  species,  Enphysora  bigclowi  Maas,  to  the  genus 
(  orymorpha  within  the  subgenus  Enphysa;  and  Mayer  (1910)  combined  Euphy- 
sora and  Enphysa  into  Corymopha.  Vanhoffen  (1911)  and  Browne  (1916) 
thereupon  retained  Enphysora.  but  little  more  than  a  decade  later  Uchida  (1927) 
assigned  E.  bigchu'i  to  Enphysa.  The  following  year  Kramp  (1928)  argued  for 
the  retention  of  Enphysora.  further  suggesting  that  the  genus  was  more  closely 
allied  to  Corymorpha  than  to  Enphysa.  While  demonstrating  the  distinctness  of 
the  polyp  stages  of  Enphysa  and  Corymorpha.  Kramp  (1949)  did  not  speculate 
further  on  the  position  of  Euphysora. 

The  ambiguity  in  the  systematic  position  of  Enphysora  arose  for  three  reasons. 
First,  the  criteria  upon  which  Maas  (1905)  based  this  new  genus  were  felt  by 
some  workers  to  be  arbitrary  and  susceptible  to  individual  interpretation  (Mayer, 
1910).  Secondly,  the  genus  as  currently  constituted  (Kramp,  1961)  may  be  a 
heterogeneous  mixture  of  species  (Kramp,  1948).  Finally,  and  most  significantly, 
the  polyp  stage  has  not  been  described  for  any  species  of  Enphysora  (Kramp, 
1961)." 

In  this  paper  the  life  cycle  of  the  type  species,  Enphysora  bigelowi,  is  described. 
On  the  basis  of  this  life  cycle  it  is  necessary  to  revise  the  nomenclature  of 
Enphysora  to  reflect  a  close  alliance  with  the  higher  Corymorphines  and  to  con- 
tinue the  recent  trend  of  elimination  of  dual  classification  in  hyclrozoans  (Rees, 
1957;  Naumov,  1960;  Rrinckmann-Voss,  1970).  The  implications  of  this  revision 
on  the  systematic  positions  of  other  members  of  the  medusa  genus  are  also 
discussed. 

MATERIALS  AND  METHODS 

Five  sexually  mature  corymorphid  medusae,  later  identified  as  Euphysora 
bigeloztn,  were  collected  in  a  plankton  tow  in  Monterey  Bay,  on  September  24, 
1973.  The  net  (0.5  m  diameter)  was  towed  on  a  weighted  15  m  line  at  very 
slow  speed  for  about  30  min,  and  was  on  the  bottom  during  part  of  the  tow 
(some  sediment  was  recovered  with  the  sample). 

The  medusae  were  returned  to  the  laboratory,  fed  brine  shrimp  nauplii,  and 
left  in  a  small  finger  bowl  at  about  14°  C  for  several  days.  They  died  shortly 
thereafter  and  disintegrated.  At  this  time,  however,  metamorphosing  larvae  were 
noticed  on  the  bottom  of  the  bowl  and  the  sea  water  was  replaced.  Within  a  week 
these  larvae  had  completed  metamorphosis  into  eight  polyps.  Several  polyps  were 
transferred  to  the  Bodega  Marine  Laboratory  for  culture  ;  the  remainder  were  main- 
tained at  Stanford. 

The  first  evidence  of  gonosome  development  in  the  Stanford  culture  was  on 
October  17,  1973,  and  the  first  medusa  was  liberated  on  December  5,  1973.  The 
Stanford  culture  deteriorated  shortly  thereafter,  and  the  line  was  lost  in  January, 
1('74.  Attempts  to  rear  young  medusae  on  brine  shrimp  nauplii  were  unsuccessful. 
The  descriptions  which  follow  are  based  on  a  preserved  polyp,  five  newly  released 
medusae  preserved  within  24  hr  of  liberation,  notes  made  on  living  polyps  and 


LIFE  CYCLE  OF  CORYMORPII.  I  inGELOWl  487 

medusae,  and  extensive  photographic  records  of  various  stages.  The  polyp  and 
five  newly  released  medusae  have  heen  deposited  at  the  National  Museum  of 
Natural  History  (#56762  and  #56760). 

RESULTS 

Adult   medusa 

An  adult  medusa  is  shown  in  Figure  Ic.  The  medusae  were  corymorphid  in 
morphology  with  three  short  simple  tentacles  and  one  long  tentacle  which  differed 
from  the  others  in  form  as  well  as  size.  All  tentacles  were  hollow.  The  length 
of  the  bell  ranged  up  to  5.0  mm.  When  extended,  the  manubrium  reached  the 
unibrellar  margin  and  sometimes  protruded  slightly  beyond.  The  edges  of  the 
mouth  were  armed  with  nematocysts.  The  apical  projection  was  produced  as  a 
conical  process  with  an  apical  canal  extending  about  two  thirds  of  the  way  to 
the  tip.  The  primary  tentacle  was  as  long  as  the  bell  and  was  studded  with  as 
many  as  eight  (possibly  nine)  subterminal  nematocyst  bulbs  along  its  length. 
The  largest  specimen  had  a  large,  club-shaped  terminal  bulb  which  may  have 
represented  a  small  terminal  bulb  and  a  ninth  subterminal  bulb  in  the  process  of 
division.  All  subterminal  nematocyst  bulbs  were  adaxial  in  orientation.  The 
remaining  three  tentacles  were  short  and  simple,  and  were  not  armed  with 
nematocyst  bulbs,  but  with  scattered  nematocysts.  The  two  tentacles  adjacent 
to  the  primary  tentacle,  the  "lateral"  tentacles  (Kramp,  1928),  were  twice  as  long 
as  the  one  opposite.  The  morphology  of  these  medusae  was  well  within  the  range 
of  variation  of  previous  descriptions  of  Euphysora  bigelowi  (Maas,  1905;  Browne, 
1916;  Uchida,  1927;  Kramp,  1928). 

Zoogeographic  records  of  Euphysoni  biyclowi  indicate  a  wide,  warm- water 
distribution.  Since  its  original  collection  in  the  Malay  Archipelago  (Maas, 
1905),  its  known  distribution  has  been  extended  to  include  the  Indian  Ocean 
(Browne,  1916),  northeastern  Australia  (Kramp,  1953),  and  southeastern  Japan 
(Uchida,  1938;  Yamazi,  1958).  It  ranges  westward  across  the  Pacific  Ocean  to 
the  Palau  Islands  (Uchida,  1947)  and  has  been  reported  from  Chile  (Kramp, 
1952).  Kramp  (1968)  later  suggested  that  the  Chilean  record  might  be  errone- 
ous, although  it  now  appears  less  suspect.  The  collection  of  Euphysora  bigelozvi 
in  Monterey  Bay,  California,  is  rather  surprising  and  is  inconsistent  with  all  pre- 
vious records  except  that  of  Chile.  The  Monterey  Bay  collection  represents  a 
substantial  range  extension  (of  about  8,000  km)  into  the  northeastern  Pacific 
Ocean. 

In  comparing  the  various  descriptions  of  Euphysora  bigclovci  with  the  Cali- 
fornia material,  it  is  clear  that  there  is  extensive  morphological  variation  in  this 
species,  both  within  and  between  populations.  Characters  which  have  proven  to 
be  quite  variable  are  the  presence  or  absence  of  the  apical  canal,  the  relative  lengths 
of  the  three  secondary  tentacles,  and  the  relationship  between  bell  height  and  the 
number  of  nematocyst  bulbs  on  the  principle  tentacle.  Populations  from  the 
Malay  Archipelago  (Maas,  1905)  included  animals  with  and  without  apical 
canals;  and  Kramp  (1928)  reported  variation  for  specimens  from  the  Sunda 
Strait  (the  predominant  type  there  was  lacking  the  canal).  Other  descriptions 


488 


C.  SASSAMAN  AND  J.  T.  KEES 


a 


FIGURE  1.  Stages  in  the  life  cycle  of  Coryinorpha  (=  Euf>h ysora)  bigeloivi:  a,  composite 
drawing  of  the  mature  polyp  illustrating  its  general  aspect  and  emphasizing  characteristic 
features;  b,  the  newly  released  medusa;  c,  the  adult  medusa;  d,  polyp  reproduction  by  frustula- 
tion  (the  fragmented  base  of  the  parental  polyp  is  shown  to  the  right  of  the  metamorphosing 
bud).  Scale  bar  is  1.0  mm  for  (a)  and  (c)  and  is  0.5  mm  for  (b)  and  (d). 


LIFE  CYCLE  OF  CORYMORPHA  BIGELOWI  489 

(e.g.,  Browne,  1916;  Uchida,  1927)  indicate  the  complete  absence  of  apical  canals 
in  animals  from  the  Indian  Ocean  and  off  Japan.  The  California  specimens,  in 
contrast,  all  had  well-developed  apical  canals.  Variation  in  the  relative  lengths 
of  the  three  secondary  tentacles  is  also  substantial.  Kramp  (1928)  tabulated  the 
relative  lengths  of  the  "lateral"  and  "opposite"  secondary  tentacles  for  the  Sunda 
Strait  specimens.  His  analysis  indicated  that  the  opposite  tentacle  is  shorter  than 
the  other  two  in  small  specimens  (1.5  mm  high),  but  that  its  relative  length 
increases  with  medusa  size  and  may  eventually  exceed  the  lateral  tentacles  in 
length  (in  2.25  to  3  mm  high  medusae).  In  contrast,  Browne  (1916)  noted  that 
in  his  small  specimens  the  three  secondary  tentacles  were  of  equivalent  length,  but 
that  in  the  larger  specimen  (4  mm  high)  the  opposite  tentacle  was  much  shorter 
than  the  lateral  tentacles.  In  the  California  specimens  the  opposite  tentacle  was 
substantially  shorter  than  the  other  two,  even  in  the  largest  (5  mm  high)  medusa 
(Fig.  Ic).  There  appears  to  be  differential  development  of  the  secondary  tentacles, 
with  variation  among  populations.  For  this  character  our  specimens  are  more  like 
those  from  the  Indian  Ocean  than  those  from  the  Sunda  Strait.  A  third 
morphological  feature  showing  substantial  variation  is  the  number  of  nematocyst 
bulbs  on  the  principal  tentacle.  This  character  seems  to  be  related  to  medusa 
height  (Browne,  1916;  Kramp,  1928),  but  the  degree  to  which  the  number  of 
bulbs  increases  per  unit  change  in  medusa  height  seems  to  vary  among  populations. 
For  example,  a  4.0  mm  high  medusa  from  the  Indian  Ocean  had  1 1  nematocyst 
bulbs  (Browne,  1916),  whereas  a  2.25  mm  medusa  from  the  Sunda  Strait  had 
21  bulbs,  and  one  individual  1.5  mm  high  had  31  (Kramp,  1928).  Uchida 
(1927)  illustrates  a  3.5  mm  medusa  with  26  subterminal  bulbs.  For  this  char- 
acter the  California  sample  is  more  similar  to  that  from  the  Indian  Ocean  than  to 
the  Sunda  Strait  or  Japanese  collections.  Variation  in  the  three  characters  does 
not  appear  to  be  correlated.  The  California  specimens  resemble  those  from  the 
Indian  Ocean  with  regard  to  the  lengths  of  the  secondary  tentacles  and  the 
number  of  nematocyst  clusters  on  the  primary  tentacle,  but  in  one  group  the  apical 
canal  was  uniformly  lacking  and  in  the  other  it  was  uniformly  well-developed. 
The  use  of  these  characters  in  delineating  the  genetic  relationships  between 
populations  in  different  parts  of  the  species  range  will  probably  not  be  very 
productive. 

Morphology  of  the  pol\p 

The  following  description  was  made  from  a  polyp  grown  in  the  laboratory  until 
preservation  on  December  12,  1973,  and  from  notes  and  photographs  of  live  polyps 
in  culture.  The  preserved  specimen  is  13  mm  high  and  about  1  mm  wide  at  its 
widest  point.  The  hypostome  is  3  mm  high  and  also  about  1  mm  wide.  There 
is  considerable  variation  in  dimensions  depending  upon  the  state  of  expansion 
in  live  individuals.  Figure  la  illustrates  the  general  aspect  of  the  polyp  and 
emphasizes  some  of  the  characteristic  morphological  features.  Figure  2a  shows 
the  hypostome  and  the  early  gonosome  of  a  mature  polyp. 

Among  individuals  there  are  between  15  and  20  aboral  filiform  tentacles  (beset 
with  scattered  nematocysts)  in  a  single  whorl.  These  tentacles  are  apparently 
not  very  contractile;  our  photographs  do  not  include  any  in  which  the  aboral 


490  C.  SASSAMAN  AND  J.  T.  REES 

tentacles  are  substantially  contracted  despite  the  use  of  intense  lighting  and  the 
occasional  addition  of  brine  shrimp  nauplii  during  photographing.  In  mature 
polyps  there  are  up  to  35  oral  tentacles  (with  scattered  nematocyst  batteries)  set 
in  irregular  rows  on  the  hypostome.  Although  the  oral  tentacles  are  not  distinctly 
capitate,  they  may  be  somewhat  thickened  at  their  tips,  particularly  in  young  polyps. 
A  diaphragm  separates  the  hypostome  from  the  polyp  body  (Fig.  la).  The  hydro- 
caulus  is  enclosed  in  a  thin,  membranous  perisarc  which  is  attached  to  an  annular 
ring  of  thickened  ectoderm  slightly  below  the  diaphragm.  In  some  specimens  the 
perisarc  extended  beyond  the  base  of  the  hydranth  in  the  form  of  a  thin  tube. 
The  body  is  slightly  inflated  at  its  base  to  a  width  of  about  1.5  mm.  Anchoring 
rootlets  with  inflated  tips  and  varying  in  width  between  25  and  50  /xin  arise  from 
prominent  endodermal  canals  which  are  visible  in  the  hydrocaulus  (Fig.  la).  The 
medusa  buds  are  mounted  in  clusters  on  inflated  pedicels  which  arise  from  the 
hypostome  between  the  oral  and  aboral  tentacles,  but  much  nearer  the  aboral 
tentacles  (Fig.  2a).  These  pedicels  are  not  very  long  (1  to  2  mm)  and  are  not 
highly  branched  (Fig.  2b). 

Gonosoinc  development  and   tlie  ne?\.'l\  liberated   medusa 

Gonosotne  development  was  first  observed  about  two  weeks  after  larval 
metamorphosis.  Subsequent  development  of  the  medusa  buds  was  substantially 
slower  than  initiation  of  the  gonosome,  and  the  first  medusa  was  not  released 
until  about  six  weeks  after  the  gonodendra  were  first  visible. 

The  gonodendra  develop  asynchronously  on  the  polyp,  several  stages  of  pro- 
gression being  found  on  the  same  hydranth.  The  earliest  structure  is  a  simple 
tubular  projection  from  the  hypostome  immediately  above  the  aboral  tentacles. 
This  projection  elongates  and  branches,  the  medusa  buds  forming  at  the  termini 
of  each  branch  (Fig.  2b).  The  differential  development  of  the  primary  tentacle 
of  the  medusa  takes  place  during  attachment  to  the  gonodendra  and  the  manubrium 
swells  to  occupy  most  of  the  subumbrellar  cavity  (Fig.  2c).  Approximately  a  day 
before  liberation  the  attached  medusa  falls  below  the  whorl  of  aboral  tentacles 
(Fig.  2d)  by  which  time  the  medusa  is  contractile,  but  is  not  rhythmically  pulsating. 
It  is  released  (or  breaks  free)  with  an  incompletely  formed  apical  chamber  at  a 
size  of  about  1.3  mm  high  by  1.2  mm  wide. 

The  newly  released  medusa  (Fig.  Ib)  is  colorless  except  for  pale  yellow 
tentacle  bulbs.  The  manubrium  is  tubular  and  extends  to  the  velar  opening  or 
slightly  beyond.  The  apical  canal  is  variable  in  development,  extending  from 
one-sixth  to  two-thirds  of  the  way  to  the  tip  of  the  apical  projection.  This  range 
of  variation  is  found  even  between  individual  medusae  released  from  the  same 
polyp.  The  tip  of  the  apical  projection  has  small  papillae  on  its  surface.  Only 
one  tentacle  is  developed  to  any  appreciable  degree,  the  others  being  reduced  to 
conical  projections.  The  primary  tentacle  bears  a  club-shaped  terminal  nemato- 
cyst bulb,  but  is  lacking  the  subterminal  adaxial  bulbs  of  older  medusae.  Nemato- 
cysts  are  present  on  all  tentacles,  but  are  lacking  on  the  exumbrellar  surface. 

The  asynchronous  development  of  the  gonosome  results  in  a  prolonged  period 
of  medusa  liberation.  One  polyp  was  censused  daily  and  produced  a  total  of 
34  medusae  over  a  period  of  three  weeks.  Newly  released  medusae  were  not 


LIFE  CYCLE  OF  CORYMORPHA  BIGELOWI 


491 


seen  to  feed  on  brine  shrimp  nauplii  and  did  not  live  for  more  than  a  few  days. 
The  newly  released  medusae  are  similar  in  size  to  the  smaller  individuals 
described  from  plankton  collections  by  Kramp  (1928)  and  Browne  (1916).  It  is 
of  interest  to  note  that  in  these  collections  small  individuals  (1.25  to  1.5  mm 
high)  have  subterminal  nematocyst  bulbs  on  the  primary  tentacle  which  are 
absent  on  the  newly  released  Euphysora  bigelozvi  medusae.  In  Coryinorpha  nutans 
the  annular  subterminal  bulbs  are  developed  even  before  liberation  from  the  polyp 
(Russell,  1953). 


FIGURE  2.  The  development  of  the  gonosome  of  Corymorpha  (=  Euphysora)  bigclozvi: 
a,  lateral  view  of  the  hypostome  of  a  polyp  with  early  gonodendra ;  b,  immature  gonophore  with 
inflated  pedicel  and  developing  medusa  buds ;  c,  medusa  buds  in  advanced  state  of  development 
(note  the  enlarged  manubrium  and  primary  tentacle)  ;  d,  medusa  just  prior  to  release. 
Scale  bar  is  0.5  mm. 


C.  SASSAMAN  AND  J.  T.  REES 


TABLE  I 
Cniihnu  o/  Corymorpha  (  =  Euphysora)  bigelowi,  with 


n  >u<:nni.\. 


Stage 

Stenoteles 
(large) 

Stenoteles 
(small) 

Microbasic 
mastigophores 

Desmonemes 

Anisorhizas* 

Polyp 

Oral  and  aboral 

tentacles 

13-16 

7-8  X  4.5-5 

7.5-8.5 

4-6 

X  8.5-10 

X  3-4 

X  3.5-4.5 

\r\vly  released  medusa 

Tentacles 

11-15 

8-10  X  7-8 

4.5  X  9.5 

6.5-9 

X  9-12 

X  3.5-5.5 

Adult  medusa** 

Primary  tentacle 

13-14 

7.5  X  7 

c  10 

X  11-12 

Secondary  tentacle 

11  X  13 

8-9  X  7 

11  X  12 

Lips  of  mouth 

12-14 

6  X  7 

X  10-12 

Umbrella 

7  X  9 

3X8 

9  X  10 

*  No  fired  nematocysts  of  this  type  were  closely  examined. 

*  Measurements  from  photographs  of  tissue  squashes. 

Cnidom 

Data  on  the  sizes,  types,  and  locations  of  nematocysts  present  in  different 
stages  of  the  life  cycle  are  given  in  Table  I,  and  selected  types  are  illustrated 
in  Figure  3.  Four  types  of  nematocysts  were  found :  Stenoteles,  microbasic 
mastigophores,  desmonemes,  and  probably  anisorhizas.  These  four  types  have 
been  previously  reported  in  two  related  species,  Corymorpha  nutans  Sars  and 
Ectopleura  dnmortieri  (Van  Beneclen)  (Russell,  1938;  Weill,  1934).  In  Enphysa 
aurata  Forbes  the  heteronemes  and  desmonemes  have  not  been  found  and  atrichous 
haplonemes  are  present  (Rees,  1957). 

Mode  of  polyp  asexual  reproduction 

Asexual  polyp  reproduction  was  observed  once.  The  terminal  portion  of  the 
base  of  the  polyp  detached  from  the  remainder  of  the  hydrocaulus,  and  within  a 
few  days  this  fragment  began  development  of  both  oral  and  aboral  tentacles  and 
broke  free  of  the  parental  perisarc  (Fig.  Id).  This  mode  of  reproduction  has  been 
termed  "frustulation"  by  Kramp  (1948)  to  emphasize  its  relationship  to  transverse 
fission  and  its  distinctness  from  true  budding.  It  is  apparently  a  normal  process  of 
reproduction  in  several  species  of  Enphysa  (e.g.,  Miles,  1937)  but  has  not  been 
reported  for  Corymorpha  (Kramp,  1949)  in  which  other  forms  of  asexual  re- 
production, such  as  polyp  development  from  root  filaments,  are  observed  (Ikeda, 
1910;  Kramp,  1949). 

DISCUSSION 


The  polyp  reared  from  Euphysora  bigelowi  demonstrates  the  following  struc- 
tural features  which  are  characteristic  of  the  genus  Corymorpha:   an  irregularly 


LIFE  CYCLE  OF  CORYMORPHA  BIGELOWI 


493 


arranged  cluster  of  oral  tentacles,  a  single  whorl  of  aboral  tentacles  with  scattered 
(as  opposed  to  annular)  nematocyst  batteries,  an  annular  diaphragm,  rooting 
filaments  borne  only  in  the  lowermost  part  of  the  hydrocaulus,  gonophores  borne  in 
clusters  on  pedicels  (gonodendra),  and  well  developed  endodermal  canals  in  the 
lower  part  of  the  hydrocaulus.  Indeed,  the  parenchymous  diaphragm  and  endo- 
dermal canals  are  considered  to  be  characteristic  of  advanced  and  highly  specialized 
members  of  the  genus  (Kramp,  1949).  However,  the  polyp  of  E.  bigelozvi  also 
shares  certain  features  with  Euphysa-type.  polyps.  The  initially  somewhat  capitate 
oral  tentacles  of  E.  bigelozvi  apparently  do  not  become  completely  filiform  in  older 
polyps  and  are  intermediate  between  Corymorpha  and  Euphysa  in  this  regard. 
The  curious  mode  of  asexual  reproduction  by  frustulation  (Fig.  Id)  has  been 
reported  in  Euphysa  (Miles,  1937)  but  not  in  Corymorpha  (Kramp,  1949).  The 
perisarc  of  E.  bigelozvi  is  reminiscent  of  Euphysa  both  in  its  attachment  to  the 
upper  region  of  the  hydrocaulus  and  in  its  extension  beyond  the  base  of  the  hydro- 
caulus. We  do  not  know  the  extent  to  which  peculiar  environmental  factors  of 
our  culture  conditions  may  have  influenced  the  expression  of  these  characteristics, 
or  the  degree  of  phenotypic  plasticity  in  this  species.  Modes  of  budding  in  hydro- 


d 


FIGURE  3.  The  cnidom  of  Corymorpha  (==  Euphysora)  bigelozvi:  from  aboral  tentacle 
of  polyp  (a-c)- — a,  large  stenotele,  b,  small  stenotele,  c,  undischarged  and  discharged  desmoneme ; 
d,  undischarged  and  discharged  microbasic  mastigophore  from  oral  tentacle  of  polyp;  e, 
anisorhiza  (?)  from  secondary  tentacle  of  adult  medusa.  Scale  bar  is  10  n. 


4()4  C.  SASSAMAN  AND  J.  T.  REES 

zoans  are  quite  variable,  and  bizarre  forms  can  be  produced  under  unnatural  culture 
conditions   (Sassaman,   1974). 

The  substantial  similarities  of  the  reared  polyp  to  Corymorpha  far  outweigh 
in  significance  the  minor  deviations  from  the  typical  form;  thus,  the  polyp  can  be 
relegated  to  the  genus  Corymorpha.  Corymorpha  nutans  is  the  only  other  meta- 
genic  polyp  in  the  genus  whose  medusa  is  known.  Following  recent  efforts  in 
eliminating  the  dual  classification  system  which  has  bedeviled  hydrozoan  syste- 
matics  from  its  inception  (Rees,  1957;  Naumov,  1960;  Brinckmann-Voss,  1970), 
it  is  deemed  appropriate  to  refer  to  both  the  polyp  and  the  Euphysora  bigchzvi 
medusa  as  Corymorpha  bigclowi,  since  Corymorpha  (Sars,  1835)  precedes 
Euphysora  (Maas,  1905). 

Since  this  revision  is  based  on  the  type  species  of  the  genus  Euphysora,  and 
the  genus  is  believed  to  be  a  heterogeneous  mixture  of  species  (Kramp,  1948), 
other  medusae  previously  assigned  to  Euphysora  are  of  an  uncertain  status.  These 
species  are  E.  gracilis  (Brooks,  1882),  E.  annulata  (Kramp,  1928),  E.  furcata 
(Kramp,  1948),  E.  gigantea  (Kramp,  1957),  E.  normani  (Browne,  1916),  and 
E.  valdiviae  (Vanhoffen,  1911).  This  assemblage  includes  species  which  resemble 
C.  bigchivi  in  having  unbranched  primary  tentacles  (E.  annulata  and  E.  gracilis}, 
species  with  branched  primary  tentacles  which  lack  subterminal  nematocyst  bulbs 
(E.  furcata,  E.  gigantea,  and  E.  valdiviae},  and  two  species  (E.  valdiviae  and 
E.  normani)  with  exumbrellar  nematocyst  tracts.  This  latter  condition  may  be  a 
more  primitive  condition  than  is  typical  in  Corymorpha  (Rees,  1957).  Ultimate 
resolution  of  the  systematic  positions  of  these  various  species  will  require  addi- 
tional life  cycle  data. 

It  is  uncertain  whether  or  not  the  polyp  of  C.  bigelowi  has  been  found  in 
nature.  No  polyps  similar  to  C.  bigelo^vi  are  known  from  central  California. 
Extensive  hydroid  collections  by  the  Allan  Hancock  Foundation  Expeditions 
(Fraser,  1948)  have  not  yielded  any  local  metagenic  Corymorpha,  and  the  Pacific 
fauna,  in  general,  includes  few  metagenic  Corymorpha  species.  Uchida  (1927) 
suggested  either  C.  tomocnsis  Ikeda  or  C.  cornea  (Clark)  as  the  polyp  stage  of 
Euphysora  bigelozvi.  The  morphology  of  C.  tomocnsis  (Ikeda,  1910)  is  similar 
to  that  of  C.  bigcloivi,  particularly  the  medusa  buds.  There  are,  however,  sub- 
stantial differences  in  hydranth  size,  number  and  morphology  of  the  tentacles, 
complexity  and  development  of  the  basal  region,  and  mode  of  budding.  In 
addition,  C.  tomoensis  has  not  been  reported  from  western  North  America 
(Fraser,  1948).  Corymorpha  cornea  (Clark,  1876),  while  reported  from  North 
America,  has  not  been  adequately  described,  and  its  known  distribution  is  restricted 
to  northern  Alaska  (Torrey,  1902).  At  present,  C.  bigelowi  cannot  be  posi- 
tively associated  with  any  other  previously  described  species  of  Corymorpha, 
although  C.  tomocnsis  and  C.  cornea  cannot  be  unequivocally  eliminated.  It  is 
possible  that  the  polyp  phase  of  C.  bigelowi  has  not  yet  been  found  in  the  field. 

This  study  was  supported  by  a  Predoctoral  Fellowship  from  the  National 
Science  Foundation.  We  thank  Glenn  Drewes  for  preparing  the  illustrations, 
Dr.  Cadet  Hand  for  his  taxonomic  advice,  and  Dr.  L.  R.  G.  Snyder  for  rowing 
the  boat. 


LIFE  CYCLE  OF  CORYMORPHA  BIGELOWl  495 

SUMMARY 

1.  Five  individuals  of  the  corymorphid   jellyfish,   Euphysora   bigelowi   Maas, 
were  collected  in  1973  in  Monterey  Bay,  California,  for  a  range  extension  of  more 
than  8,000  km  across  the  northeastern  Pacific  Ocean. 

2.  Larvae  released  by  these  medusae  were  cultured  and  the  resulting  polyps, 
the  first  known  from  this  medusa  genus,  are  described. 

3.  The  polyps  are  a  Corymorpha,  but  share  some  minor  characteristics  with 
polyps  of  the  corymorphid  genus  Enphysa. 

4.  The  polyp  and  medusa  are  assigned  the  name  Corymorpha  bigcloun  (Maas)  ; 
the  systematic  implications  of  this  revision  are  discussed. 

LITERATURE  CITED 

ALLMAN,   G.  J.,   1871-1872.     A  monograph  of  the  gymnoblastic  or  tubularian  hydroids.     Ray 

Society,  London,  450  pp. 

BRINCKMANN-VOSS,  A.,   1970.     Anthomedusae/Athecate    (Hydrozoa,   Cnidaria)    of  the   Medi- 
terranean.    Fauna  Flora  Golfo  Napoli  Monogr.,  39  :   1-107. 
BROOKS,  W.  K.,  1882.    List  of  medusae  found  at  Beaufort,  N.  C,  during  the  summers  of  1880 

and  1881.     Johns  Hopkins  Univ.  Stud.  Biol.  Lab.,  2:  135-146. 
BROWNE,   E.  T.,   1916.     Medusae  from  the   Indian  Ocean.     Trans.   Linn.  Soc.  London  Zool., 

17:  169-210. 
CLARK,  S.  F.,  1876.     Report  of  the  hydroids  on  the  coast  of  Alaska  and  the  Aleutian  islands 

collected   by   W.    D.    Ball,    from    1871    to    1874.     Proc.   Acad.   Nat.   Sci.   Phila.,   28: 

205-238. 
FORBES,  E.,   1846.     On  the  pulmograde  medusae  of  the  British  seas.     Ann.  Mag.  Nat.  Hist., 

So:  1,  18  :  284-287. 
FORBES,  E.,   1848.     A  monograph   of  the  British  naked-eye  medusae.     Ray   Society,   London, 

104  pp. 
FRASER,  C.  M.,  1948.     Hydroids  of  the  Allan  Hancock  Pacific  Expeditions  since  March,  1938. 

Allan  Hancock  Pac.  Expcd.,  4:   179-335. 
HAECKEL,  E.,  1879.    Das  system  der  meduscn.    Jena,  360  pp. 
HARTLAUB,   C.,    1907.     Craspedote   Medusen,    Tiel.    I,   Lief.    I.   Codoniden   und   Cladonemiden. 

Nor.  Plankton,,  6  :  1-135. 

IKEDA,  J.,  1910.    On  a  new  species  of  Corymorpha  from  Japan.    Annot.  Zool.  Jpn.,  7  :  153-165. 
KRAMP,  P.  L.,  1928.     Papers  from  Dr.  fh.  Mortenson's  Pacific  Expedition  1914-16.     XLIII. 

Hydromedusae.     I.     Anthomedusae.      Vidcnsk.  Mcdd.  Dan.   Naturhist.   Forcn,   Kbh., 

85  :  27-64. 
KRAMP,    P.    L.,    1948.      Trachymedusae    and    Narcomedusae    from   the    'Michael    Sars'    North 

Atlantic  Deep-Sea  Expedition   1910  with   additions  on  Anthomedusae,   Leptomedusae 

and  Scyphomedusae.     Rep.  Sars  N.  All.  Deep-Sea  Expcd.  1910,  5:   1-23. 
KRAMP,   P.   L.,   1949.     Origin  of  the  hydroid  family   Corymorphidae.     Vidensk.  Mcdd.   Dan. 

Naturhist.  Forcn.  Kbh.,   Ill:   183-215. 
KRAMP,  P.  L.,  1952.    Reports  on  the  Lund  University  Chile  Expedition,  1948-49.     2.     Medusae 

collected  by  the  L.  U.  Exp.  1948-49.    Ada  Univ.  Lund.,  N.  F.  Avd.,  2,  47 :  1-19. 
KRAMP,  P.  L.,  1953.     Hydromedusae.     Sci.  Rep.  Great  Barrier  Reef  Exped.,  6 :  259-322. 
KRAMP,  P.  L.,  1957.     Hydromedusae  of  the  Discovery  Collections.    Discovery  Rep.,  29 :  1-128. 
KRAMP,  P.  L.,  1961.     Synopsis  of  the  medusae  of  the  world.     /.  Mar.  Biol.  Assoc.  U.K.,  40  : 

1-469. 
KRAMP,  P.  L.,  1968.     The  hydromedusae  of  the  Pacific  and  Indian  Oceans.     Sections  II  and 

III.    Dana-Rep.  Carlsberg  Found.,  72  :  1-200. 
MAAS,  O.,  1905.     Die  Craspedoten  Medusen  der  Siboga  Expedition.     Siboga  Expcd.  Monogr., 

10:  1-85. 

MAYER,  A.  G.,  1910.    Medusae  of  the  world.     The  hydromedusae,  Vols.  I,  II.     Carnegie  Insti- 
tution, Washington,  498  pp. 


496  C.  SASSAMAN  AND  J.  T.  REKS 

MILES,  S.  S.,   1937.     A  new  genus  of  hydroid  and  its  method  of  asexual  reproduction.     Biol. 

Bull.,  72:  327-333. 
NAUMOV,  D.  V.,  1960.     Hydroids  and  hydromedusae  of  the  USSR.     Academy  of  Sciences  of 

the  USSR,  Moscow,  660  pp. 
REES,  W.  J.,  1957.     Evolutionary  trends  in  the  classification  of  capitate  hydroids  and  medusae. 

Bull.  British  Mus.   (Nat.  Hist.)   Zool,  4:  455-534. 
RUSSELL,  F.  S.,  1938.     On  the  nematocysts  of  hydromedusae.     /.  Mar.  Biol.  Assoc.  U.K.,  23  : 

145-165. 

RUSSELL,  F.  S.,  1953.     Medusae  of  the  British  Isles.     Vol.  I,  Hydromedusae.     Cambridge  Uni- 
versity Press,  530  pp. 
SARS,    M.,    1835.      Beskrivclser    og    lagttagclscr    over   nogle    macrkeliye    cllcr    nye    i    Plavet 

ved  den  Bergcnske  Kyst  levcnde  Dyr.     Bergen,  81  pp. 
SASSAMAN,  C.,   1974.     Capacity  for  development  of  secondary  manubria  in  Eutonina  indicans 

medusae  (Hydrozoa).    Pac.  Sci.,  28:  375-376. 
TORREY,  H.  B.,   1902.     The   Hydroida  of  the   Pacific  Coast  of   North  America.      Univ.   Calif. 

Publ.  Zool,  1:   1-104. 
UCHIDA,  T.,  1927.    Studies  on  Japanese  hydromedusae.     I.    Anthomedusae.    /.  Fac.  Sci.  Tok\o 

Univ.,  1:   145-241. 
UCHIDA,  T.,  1938.     Medusae  in  the  vicinity  of  the  Amakusa  Marine  Biological  Station.     Bull. 

Biogeogr.  Soc.  Jpn.,  8:   143-149. 
UCHIDA,  T.,   1947.     Some  medusae  from  the  central   Pacific.     /.  Far.  Sci.  Hokkaido    Univ., 

Ser.  6,  Zool,  9 :  297-319. 
VANHOFFEN,  E.,  1911.     Die  Anthomedusen  und  Leptomedusen  der  Deutschen  Tiefsee-Expedi- 

tion  1898-1899.     Wiss.  Ergebn.  Valdivia,  19:  193-233. 
WEILL,   R.,    1934.     Contribution   a   1'etude  des   Cnidaires   et   leurs   nematocysts.      II.     Valeur 

taxonomique  du  cnidome.     Trav.  Stat.  Zool   IVimcreux,   11  :   349-701. 
YAMAZI,   I.,   1958.     Preliminary  check-list   of  plankton   organisms   found   in   Tanabe   Bay   and 

its  environs.     Publ.  Seto  Mar.  Biol.  Lab.,  7:  111-163. 


Reference:  Biol.  Bull,  154:  497-507.     (June,  1978) 


THE  ANATOMY  OF  THE  DECAPOD  CRUSTACEAN 

AUXILIARY  HEART 

A.  STEINACKERi 

Department  of  Biology,  University  of  California,  San  Diego  92093;  and 
Department  of  Biology,  Stanford   University,  Stanford,  California  94305 

An  auxiliary  heart  is  found  in  many  decapod  crustaceans  at  the  anterior  end  of 
the  dorsal  median  artery  before  the  artery  branches  to  supply  the  supraesophageal 
ganglion  and  the  peripheral  oculomotor  and  visual  systems.  Although  the  existence 
of  this  auxiliary  heart  had  been  noted  earlier,  when  it  was  named  the  cor  frontale 
(Baumann,  1917),  very  little  information  was  provided  beyond  a  description  of 
the  muscles  involved  (for  review,  see  Maynard,  1960).  The  following  is  a  more 
thorough  account  of  the  anatomy  of  the  cor  frontale  in  several  decapods  with  par- 
ticular emphasis  on  the  neural  elements  of  the  system. 

MATERIALS  AND  METHODS 

Specimens  of  Callincctes  sapidiis,  the  American  blue  crab,  and  Pamtlirns 
interruptus,  the  California!!  lobster,  were  used  for  the  most  complete  dissections. 
Specimens  of  Scylla  serrata,  the  Australian  mud  crab,  and  two  Californian  marine 
crabs,  Cancer  productus  and  Cancer  antcnnarius,  were  also  investigated. 

The  primary  method  used  to  trace  the  neural  elements  was  in  vivo  methylene 
blue  staining  by  perfusion  through  the  dorsal  medial  artery.  Fixation  of  the 
material  so  stained  was  done  by  the  method  of  Pantin  (1969).  Light  microscopic 
histological  preparations  of  the  heart  nerves  and  tendon  ganglia  were  made  using 
glutaraldehyde  fixation,  Epon  embedding  and  toluidine  blue  staining.  Electron 
microscopy  of  the  muscles  was  done  with  a  3%  glutaraldehyde,  \%  paraformalde- 
hyde,  collidine  buffer,  1178  m  osmol  fixation  and  Epon  embedding.  To  trace  the 
course  of  vessels  of  the  system,  liquid  latex  (Connecticut  Valley  Biological  Supply) 
was  injected  into  the  cerebral  vascular  system  via  the  dorsal  median  artery. 

RESULTS 

The  basic  anatomy  of  the  cor  frontale,  which  is  remarkably  similar  in  all  the 
decapods  studied,  is  illustrated  in  Figures  1,  2,  and  3.  The  blood  flows  anteriorly 
from  the  main  heart  through  the  dorsal  median  artery  to  supply  the  supraesophageal 
ganglion  and  the  peripheral  optic  ganglia  and  oculomotor  system  (Fig.  1).  Before 
the  blood  is  distributed  to  these  areas,  it  flows  through  the  auxiliary  heart.  The 
anatomy  of  this  heart  can  most  conveniently  be  described  by  breaking  it  down  into 
three  elements :  the  blood  vessels,  the  muscles  and  tendons,  and  the  associated 
neural  system. 

1  Present  Address :  Department  of  Pharmacology,  College  of  Medicine  and  Dentistry  of 
New  Jersey-Rutgers  Medical  School,  Piscataway,  New  Jersey  08854. 

497 


498 


A.  STEINACKER 


OA  MH 

DTCF      S  DMA 


SOG 


VT 


FIGURE  1.  Overview  of  the  location  of  the  cor  frontale  (CF)  in  relation  to  other  struc- 
tures in  the  cephalothorax  of  the  spiny  lobster,  PanuUrus  intcrruptus.  Blood  flows  from  the 
main  heart  (MH)  through  the  dorsal  median  artery  (DMA)  over  the  stomach  (S)  to  the 
cor  frontale  (CF)  from  which  it  exits  to  the  eyecup  via  the  ophthalmic  artery  (OA)  and  to 
the  supraesophageal  ganglion  (SOG)  via  the  cerebral  artery  (CA).  Other  abbreviations 
are:  dorsal  tendons  of  the  cor  frontale  (DT)  ;  single  ventral  tendon  (VT)  ;  and  circum- 
esophageal  connectives  (CC). 

Blood  vessels 

The  wall  of  the  auxiliary  heart  is  formed  by  the  dilated  terminal  end  of  the 
dorsal  median  artery.  Note  (Figs.  2  and  3)  that  no  muscle  is  contained  in  the 
wall  of  the  artery  itself.  Rather,  the  wall  of  the  cor  frontale  is  composed  of  the 
same  two  layers  as  that  of  the  dorsal  median  artery. 

The  course  of  the  blood  vessels  from  the  cor  frontale  can  be  seen  in  the  latex 
injected  preparation  of  Callinectes  in  Figure  4.  Blood  enters  the  cor  frontale  sinus 
from  the  dorsal  median  artery  and  leaves  via  the  cerebral  artery  which  supplies  the 
supraesophageal  ganglion,  a  few  small  vessels  which  supply  the  nearby  eyestalk 
muscles  and  the  two  large  ophthalmic  arteries  which  supply  the  visual  and  oculo- 
motor system  in  the  eyecup.  (The  dorsal  median  artery  is  sometimes  referred 
to  as  the  ophthalmic  artery,  a  misnomer,  since  the  true  ophthalmic  arteries, 
which  run  to  the  eyes,  receive  only  part  of  the  supply  of  the  dorsal  median  artery.) 

Muscles  and  tendons 

The  cor  frontale  muscles  are  two  distinct  strips  of  striated  muscle  originating 
from  tendons  outside  the  dorsal  median  artery.  In  the  crab,  the  tendons  begin 
as  multiple  insertions  on  the  dorsal  carapace  just  behind  the  middle  cylinder  of 
the  eyestalk.  This  origin  can  be  seen  as  two  indentations  on  both  the  underside 
and  on  the  external  surface  of  the  dorsal  carapace.  In  the  lobster,  which  has  no 


DECAPOD  CRUSTACEAN  AUXILIARY  HEART 


499 


eyestalk  middle  cylinder,  the  tendons  originate  in  an  equivalent  position  on  the 
dorsal  carapace  between  the  two  large  spines  of  the  rostrum.  Each  of  these  two 
tendons  is  joined  by  an  orthogonal  lateral  tendon  before  the  tendons  pass  through 
the  wall  of  the  dorsal  median  artery.  At  this  point,  the  cor  frontale  sinus  begins 
(Fig.  2).  As  the  tendons  pass  into  the  sinus  they  give  rise  to  the  two  muscles 


DCFT 


DMA 


MHVN 


IVN 


VT 


SGN 


FIGURE  2.  Transverse  view  of  the  cor  frontale  of  Panulirus  intcrniptns.  Abbreviations 
are:  dorsal  cor  frontale  tendons  (DCFT);  anterior  gastric  muscles  (AGM)  ;  cor  frontale 
nerve  (CFN)  ;  alternate  course  of  the  cor  frontale  nerve  (AC)  ;  tegumentary  nerve  (TN)  ; 
stomatogastric  nerve  (SGN)  ;  ventral  tendon  (VT)  ;  inferior  ventricular  nerve  (IVN)  ; 
posterior  eyestalk  muscles  (PEM)  ;  supraesophageal  ganglion  (SOG)  ;  cor  frontale  muscles 
(CFM)  ;  cor  frontale  sinus  wall  (CFSW)  ;  ophthalmic  artery  (OA)  ;  tendon  ganglion 
(TG)  ;  occasional  separate  tendon  sensory  supply  (TSS)  ;  and  nerve  to  main  heart  valve 
(MHVN);  dorsal  median  artery  (DMA). 


500 


A.  STEINACKER 


SOG 


B 


SOG 


SGN 


AP 


FIGURE  3.  Transverse  view  of  the  cor  frontale  in  cephalothorax  of  the  Callincctes 
sapidns.  A  shows  relation  of  cor  frontale  (CF)  to  cephalic  structures.  DMA  indicates 
dorsal  median  artery  entering  the  cor  frontale.  Muscles  bordering  cor  frontale  (EM)  are 
the  eyestalk  muscles.  The  supraesophageal  ganglion  (SOG)  lies  under  the  cor  frontale 
and  receives  its  blood  supply  via  the  cerebral  artery  (CA).  B  shows  enlarged  view  of 
center  of  (A)  showing  the  cor  frontale  sinus  walls  opened  at  arrows  to  expose  the  enclosed 
cor  frontale  muscles  (CFM)  and  stomatogastric  ganglion  (SG).  The  stomatogastric  ganglion 
(SG)  lies  inside  the  cor  frontale.  SGN  is  the  stomatogastric  nerve  exiting  the  cor  frontale 
at  the  point  where  the  tendon  has  been  detached  by  dissection  from  its  apodeme  (AP). 

of  the  cor  frontale.  These  muscles  extend  the  length  of  the  sinus  and  exit  at 
the  ventral  posterior  end  as  a  single  tendon  attached  to  an  apodeme.  This 
apodeme  (Fig.  3),  which  arises  from  an  epistome  above  the  mouth,  is  a  common 
attachment  for  the  cor  frontale  muscles,  the  dorsal  eyestalk  muscles  and  several 
esophageal  muscles. 

The  cor  frontale  of  the  lobster  differs  from  that  of  the  crab  primarily  in  the 
extent  of  the  development  of  the  tendons.  The  dorsal  tendons  of  the  lobster 
are  much  larger  and  the  single  ventral  tendon  is  greatly  elongated.  These  dif- 
ferences are  consistent  with  both  the  larger  size  and  the  dor  so-ventral  elongation 


DECAPOD  CRUSTACEAN  AUXILIARY  HEART  501 

of  the  lobster  cephalothorax  (Fig.  1).  The  size  of  the  tendons  is  particularly 
striking  in  view  of  the  relatively  small  size  of  the  cor  frontale  muscles.  Associated 
with  these  tendons  is  a  well  developed  sensory  innervation. 

There  are  around  one  hundred  individual  muscle  fibers  in  a  cross  section  of 
the  muscle  (Fig.  5).  At  two  points  on  the  muscle  perimeter  are  areas  which 
contain  much  connective  tissue,  large  motor  axons  and  some  fibers  which  contain 
dense  granules  also  seen  in  the  tendon  ganglia.  The  muscles  of  the  cor  frontale 
are  striking  in  their  compactness  and  white  hue,  being  distinctly  whiter  and  more 
dense  than  the  fastest  portions  of  the  eyestalk  muscles  which  border  them.  The 
preliminary  electron  microscopy  which  was  done  shows  only  a  few  small 
mitochondria  which  may  account  for  the  whiteness  of  the  muscle.  The  banding 
pattern  of  the  sarcomeres  is  not  well  defined.  The  Z  band  is  moderately  dense 
and  appears  to  be  continuous  across  the  sarcomere.  The  sarcoplasmic  reticulum 
is  scarce  and  connections  with  the  well-developed  T  tubular  system  are  rare. 
The  appearance  is  that  of  a  crustacean  somatic,  rather  than  heart,  muscle;  and, 
in  fact,  it  has  been  suggested  that  the  cor  frontale  muscles  are  somatic  muscles 
secondarily  adapted  for  cardiac  function  (Maynard,  1960). 

Neural  anatomy 

Cor  frontale  nerve.  This  nerve,  which  is  the  neural  connection  between  the 
auxiliary  heart  and  the  supraesophegeal  ganglion,  exits  from  the  supraesophageal 
ganglion  with  the  tegumentary  nerve  and  splits  off  as  a  small  diameter  branch  to 
curve  back  and  up  to  the  dorsal  aspect  of  the  cor  frontale.  The  nerve  passes 
under  the  cor  frontale  tendon  to  enter  the  dorsal  median  artery  near  the  entry 
of  the  tendon. 

In  the  lobster,  the  cor  frontale  nerve  either  leaves  the  tegumentary  nerve 
close  to  the  ganglion  and  takes  a  direct  route  to  the  heart  (dashed  line  in  Fig.  2), 
or  more  commonly,  it  continues  with  the  tegumentary  nerve  up  to  the  anterior 
gastric  muscle  where  it  leaves  the  tegumentary  nerve  and  passes  around  the  gas- 
tric muscle  to  enter  the  dorsal  aspect  of  the  artery.  In  the  crab,  the  course  of 
the  nerve  through  the  cephalothorax  is  invariant  but  quite  long  and  difficult  to 
trace.  It  leaves  an  anterior  branch  of  the  tegumentary  nerve  laterally  and  curves 
back  to  the  cor  frontale  as  a  fine  nerve  embedded  in  the  dorsal  hypodermis. 

Cross  section  of  the  cor  frontale  nerve  shows  seven  fibers.  After  methylene 
blue  staining  one  may  observe  two  large  axons  which  can  be  traced  to  the  cor 
frontale  muscles,  one  large  axon  which  runs  out  the  dorsal  median  artery  to  the 
valve  of  the  main  heart  (see  below)  and  several  small  axons  whose  origin 
and  termination  could  not  be  determined  because  of  their  poor  staining.  The 
motor  neuron  somata  of  the  fibers  supplying  the  cor  frontale  muscles  and  main 
heart  valve  originate  in  the  supraesophageal  ganglion.  This  was  shown  by 
electrophysiological  recording  and  by  methylene  blue  staining  in  which  the  two 
fibers  can  be  traced  clearly  from  the  ganglion  to  innervate  the  muscles.  Attempts 
to  back  fill  the  somata  with  cobalt  or  procion  yellow  have,  to  date,  been  unsuccessful. 

In  the  lobster  and  crab,  motor  axons  to  the  heart  muscle  split  into  two  branches 
as  the  nerve  enters  the  artery.  One  branch  supplies  the  ipsilateral  muscle  and  the 
other  passes  across  the  artery  to  join  with  the  axons  of  the  contralateral  cor 


502 


A.  STEINACKER 


FIGURE  4.  Sketch  of  a  liquid  latex  injected  cast  of  Callincctes  sapidus  cor  frontale. 
Blood  flows  via  the  dorsal  median  artery  (DMA)  into  the  cor  frontale  (CF)  and  out  via 
the  cerebral  artery  (CA)  and  the  ophthalmic  artery  (OA).  Dorsal  and  ventral  tendons  (T) 
of  the  cor  frontale  muscles  are  shown  exiting  the  cor  frontale  sinus. 

frontale  nerve  to  supply  the  contralateral  muscle.  Since  the  same  motor  axon 
splitting  occurs  on  both  sides,  the  result  is  that  each  muscle  is  innervated  by  four 
axons,  two  from  each  half  of  the  supraesophageal  ganglion..  In  the  lobster,  this 
crossing  of  the  motor  axons  forms  a  distinct  central  neural  cross  bar  structure 
which  is  embedded  in  the  dorsal  wall  of  the  artery  near  the  tendons.  In  the  crab, 
when  the  two  cor  frontale  nerves  enter  the  artery,  they  do  not  form  the  distinct 
neural  cross  bar  structure  seen  in  the  lobster.  The  same  splitting  and  crossing 
over  of  the  motor  axons  occurs,  but  at  lower  level  closer  to  the  muscles. 

Main  heart  valve  nerve.  In  the  lobster  and  crab,  a  single  axon  from  each  cor 
frontale  nerve  joins  a  contralateral  partner  near  the  right  dorsal  tendon  to  form 
a  nerve  which  turns  away  from  the  cor  frontale  and,  embedded  in  the  arterial  wall, 
travels  the  entire  length  of  the  dorsal  median  artery.  The  two  axons  innervate 
the  valve  of  the  artery  as  the  latter  leaves  the  main  heart.  Histological  sections 
of  the  nerve  in  the  lobster  show  two  large  axons  and  three  much  smaller  axons. 
The  two  large  axons  stain  darkly  with  methylene  blue  and  clearly  go  directly  to 
and  end  in  the  arterial  valve.  The  others  stain  poorly  with  methylene  blue  and 
branch  repeatedly  in  the  arterial  wall.  The  three  small  fibers  may  be  a  sensory 
system  which  supplies  the  arterial  wall  since  the  wall  contains  no  muscle. 

The  valve  of  the  dorsal  median  artery  at  the  main  heart  has  two  components. 
One,  a  well  developed  semilunar  valve  with  cusps  opening  to  the  arterial  side 
which  appear  to  operate  passively  to  prevent  backflow  and  the  other,  a  circular 
ring  of  muscle  fibers  between  the  valve  and  the  heart  muscle  proper.  It  is  this 


DECAPOD  CRUSTACEAN  AUXILIARY  HEART  503 

circular  muscle  which  the  two  axons  innervate.  As  the  two  axons  approach  the 
valve,  each  axon  splits  first  into  two  and  then  into  four  branches.  Two  branches 
of  each  axon  charactistically  cross  the  midline  of  the  valve  so  that  each  half  of  the 
circular  muscle  ring  is  supplied  by  axons  from  both  halves  of  the  supraesophageal 
ganglion.  The  axons  terminate  in  an  extensive  plexus  among  circular  muscle 
fibers  of  the  valve. 

In  a  transilluminated  methylene  blue  stained  preparation,  these  muscle  fibers 
are  quite  distinct  from  those  of  the  main  heart  muscle  and  appear  similar  to  those 
described  by  Alexandrowicz  (1932)  in  several  crustacean  heart  valves.  From 
their  termination  on  the  circular  muscle  of  the  valve  and  the  lack  of  any  other 
nerve  supply  to  these  muscles,  it  is  assumed  that  the  axons  to  the  valve  are  motor. 

Cor  frontale  tendon  ganglia.  Tightly  adhering  to  the  upper  tendon  of  each  cor 
frontale  muscle  as  it  passes  through  the  arterial  walls  is  an  ill-defined  aggrega- 
tion of  neural  tissue  here  termed  the  tendon  ganglia  (Fig.  2).  The  motor  axons 
to  the  muscles  pass  directly  through  this  ganglion  and  cannot  be  easily  separated 
from  it.  In  light  microscopic  toluidine  blue  stained  sections  of  the  ganglion, 
several  distinct  types  of  somata  are  found.  Two  of  these  somata  types  contain 
numerous  either  large  or  small  darkly  stained  granules  suggestive  of  neurosecretory 
vesicles.  A  third  class  of  somata  exhibits  a  clear  cytoplasm  and  distinct  nucleus. 
Numerous  neural  processes  are  also  seen  in  the  ganglia,  some  of  which  contain 
the  same  granules  seen  in  the  somata.  These  processes  do  not  appear  to  form  a 
distinct  neuropile,  although  there  is  a  somewhat  circular  course  of  the  fibers  within 
the  ganglion.  The  density  of  the  ganglion  excludes  its  function  as  a  neurohaemal 
organ.  Several  types  of  fibers  enter  or  leave  the  tendon  ganglia.  There  are  many 
small  fibers  which  can  be  traced  from  the  ganglia  to  extensive  ramifications  in  the 
arterial  wall.  The  arterial  and  cor  frontale  sinus  walls  are  innervated  by  fibers 
which  appear  to  terminate  in  the  tendon  ganglia.  Also  associated  with  the  tendon 
ganglia  are  fibers  which,  when  stained  with  methylene  blue,  can  be  seen  to  branch 
extensively  over  the  upper  tendons  of  the  cor  frontale  muscles.  These  fibers 
are  presumed  to  be  sensory,  since  they  are  never  found  extending  beyond  the 
tendon  to  the  muscle.  All  these  above  fibers  are  very  fine  and  their  termination 
difficult  to  follow.  In  the  lobster,  the  tendon  sensory  fibers  sometimes  go  directly 
to  the  supraesophageal  ganglion  via  a  separate  sensory  nerve  which  joins  the  tegu- 
mentary  nerve  above  the  origin  of  the  cor  frontale  motor  nerve  (Fig.  2). 

Finally,  connections  can  sometimes  be  traced  between  the  stomatogastric  nerve 
and  the  tendon  ganglia.  In  the  lobster  the  stomatogastric  ganglion  is  located  inside 
the  dorsal  median  artery  where  this  artery  passes  over  the  stomach  (a  few 
centimeters  posterior  to  the  cor  frontale).  The  nerve  then  exits  from  the  artery 
but  adheres  to  it,  giving  off  many  small  branches,  one  of  which  can  sometimes  be 
traced  to  the  tendon  ganglia  or  to  the  nerve  carrying  the  axon  to  the:  main 
heart  valve.  In  the  crab,  the  stomatogastric  ganglion  is  enclosed  within  the  sinus 
of  the  cor  frontale  directly  between  the  two  muscles  (Fig.  3).  Two  lateral  nerves 
from  this  ganglion  leave  the  sinus  to  supply  the  anterior  gastric  muscles.  These 
gastric  nerves  leave  the  sinus  near  the  cor  frontale  tendons  and  give  off  fine 
branches  to  the  tendon  ganglia.  The  stomatogastric  ganglion  inside  the  cor 
frontale  appears  to  be  the  same  ganglion  referred  to  as  the  ventricular  ganglion 
(Baumann,  1917;  Maynard,  1960). 


504  A.  STEINACKER 

Ventral  tendon  sensory  units.  On  the  single  long  ventral  tendon  of  the 
lobster  a  distinct  sensory  innervation  is  found  which  extends  down  the  tendon. 
The  fibers  from  this  nerve  enter  the  inferior  ventricular  nerve.  The  inferior 
ventricular  nerve,  after  leaving  its  origin  in  the  supraesophageal  ganglion,  passes 
througl i  an  opening  in  the  cor  frontale  tendon.  The  sensory  units  from  the  tendon 
join  the  nerve  and  travel  toward  the  inferior  esophageal  ganglion.  In  the  crab, 
the  ventral  tendon  is  very  short  (Fig.  3),  and  the  inferior  ventricular  nerve  does 
not  pass  through  the  tendon.  No  sensory  fibers  comparable  to  those  of  the  lobster 
ventral  tendon  have  been  found  in  the  crab. 

Auxiliary  heart  in  flic  eyecup.  When  the  oculomotor  muscles  in  the  eyecups 
are  exposed  by  dissection  (while  the  ophthalmic  arteries  are  inflated  by  saline 
perfusion  or  by  liquid  latex  injection),  muscle  number  21  (nomenclature  of 
Cochran,  1935)  is  found  to  lie  within  the  arterial  lumen  and  shows  anatomical 
features  similar  to  the  cor  frontale.  This  is  most  obvious  with  the  latex  injection 
when  the  latex  is  found  within  the  lumen  of  the  artery  completely  surrounding 
the  muscle.  The  muscle  divides  easily  into  two  sections.  Part  of  the  muscle  has 
the  appearance  of  the  other  eyecup  muscles  and  the  rest  has  the  white  dense 
appearance  of  the  cor  frontale  muscles. 

The  presence  of  this  muscle  in  the  lumen  of  the  blood  vessel,  its  physical 
resemblance  to  the  cor  frontale  muscle  and  its  location  within  the  arterial  lumen 
immediately  before  the  artery  enters  the  neuropile  of  the  eyecup  indicate  that  it 
may  be  another  auxiliary  heart.  The  optic  neuropile,  like  the  supraesophageal 
ganglion,  but  unlike  most  crustacean  ganglia,  requires  a  constant  blood  supply  and 
fails  soon  after  this  flow  is  interrupted.  This  muscle  has  been  noted  before  to  be 
"heavily  vascularized"  (Sandeman,  1967),  but  its  presence  inside  the  vessel  lumen 
and  possible  auxiliary  heart  function  were  not  noted.  In  the  shrimp,  PaJacnion,  an 
eyecup  muscle  inserted  in  the  ophthalmic  artery  has  been  described  and  a  blood 
pumping  function  ascribed  to  it  (Debaisieux,  1944;  Denial,  1953). 

DISCUSSION 

The  term  auxiliary  heart,  rather  than  accessory  heart  (Maynard,  I960),  has 
been  used  here  for  the  cor  frontale,  since  its  fine  structure  and  electrophysiological 
reflex  response  (Steinacker,  1978)  suggest  a  phasic  function  which  is  recruited 
only  when  the  main  heart  activity  is  insufficient  for  the  circulatory  requirements 
of  cerebral  nervous  system.  The  anatomy  of  this  heart  reveals  a  complex  organ 
whose  function  appears  to  be  controlled  by  and  integrated  with  several  other 
systems.  From  anatomical  and  electrophysiological  evidence  (Steinacker,  1978), 
the  main  integrative  center  appears  to  be  in  the  supraesophageal  ganglion  where 
the  motoneurons  are  located.  The  tendon  ganglia  may  be  a  second,  local  integrative 
center,  with  perhaps  a  neurosecretory  function  whose  control  could  be  exerted  at 
two  sites;  as  a  direct  action  of  neurosecretory  products  on  the  muscle  and/or  by 
a  central  effect  on  the  neurons  in  the  supraesophageal  ganglion.  Since  blood  flows 
past  the  tendon  ganglia  to  the  supraesophageal  ganglion,  neurosecretory  products 
will  be  carried  directly  to  a  central  integrative  system  in  the  supraesophageal 
ganglion.  In  addition,  afferent  or  interneuronal  fibers  from  the  tendon  ganglia 
mav  travel  in  the  cor  frontale  nerves  to  or  from  the  tendon  ganglia  and  the 


DECAPOD  CRUSTACEAN  AUXILIARY  HEART 


505 


supraesophageal  ganglion.  There  are  at  least  four  fibers  in  the  cor  frontale 
nerves  which  are  not  motor  and  which  may  arise  all  or  in  part  from  the  tendon 
ganglia. 

The  involvement  of  the  stomatogastric  ganglion  with  an  auxiliary  heart 
deserves  mention.  In  a  decapod  with  an  open  venous  system  and  inflexible 
carapace,  variations  in  volume  of  a  highly  distensible  stomach  will  have  a  consider- 
able influence  on  blood  pressure.  In  addition  to  the  passive  influence  of  stomach 
volume  on  blood  pressure,  active  uptake  of  salt  and  water  by  the  gut  in  crustaceans 
has  been  demonstrated  (Weisman,  1874;  Fox,  1952;  Croghan,  1958).  The 
stomatogastric  system  may  be  involved  in  hemodynamics  through  the  passive  effect 
of  the  stomach  volume  on  blood  pressure  and  through  active  control  of  salt  and 
water  uptake.  The  stomatogastric  system  and  cor  frontale  also  have  a  possible 
neural  communication  via  the  supraesophageal  ganglion  and  lower  control  centers. 
Evidence  is  building  for  a  common  control  center  (or  centers)  for  the  gills,  main 
heart,  auxiliary  heart  and  stomatogastric  system.  Excitatory  and  inhibitory  fibers, 
which  have  been  found  in  the  circumesophageal  connectives,  govern  these  systems 
(Wiersma  and  Novitski,  1942;  Mendelson,  1971;  Field  and  Larimer,  1975; 
Wilkens,  Wilkens,  and  McAIuhon,  1974:  Steinacker,  1978).  Command  fibers 
for  the  stomatogastric  system  are  thought  to  originate  in  the  supraesophageal 
ganglion  (Dando  and  Selverston,  1972)  and  neurosecretory  cells  connecting  the 
supraesophageal  and  lower  neural  centers  exist  (Goldstone  and  Cook,  1971). 

The  sensory  innervation  of  the  tendons  of  the  cor  frontale  introduces  the 
possibility  of  either  feedback  control  or  coordination  of  cor  frontale  function  with 
the  other  cardioregulatorv  svstems.  In  the  case  of  the  ventral  tendon,  this  infor- 


:r '  4 


FIGURE  5.  Light  micrograph  of  transverse  section  of  a  single  cor  frontale  muscle  illustrat- 
ing somatic  nature  of  the  muscle.  Note  small  number  of  homogeneous  fiber  types  with  the 
exception  of  t\vo  areas  (at  arrows)  where  small  muscle  fibers,  nerve  fibers  and  connective 
tissue  stroma  is  found.  Scale  equals  100  microns. 


506  A.  STEINACKER 

niation  appears  to  be  feeding  into  the  esophageal  and/or  stomatogastric  system 
since  the  sensory  fibers  travel  away  from  the  supraesophageal  ganglion.  In  the 
dorsal  tendons,  either  the  tendon  ganglia  and/or  the  supraesophageal  ganglion 
receive  the  sensory  input  from  the  tendons.  In  addition,  the  innervation  of  the 
walls  of  the  artery  and  the  sinus  wall  of  the  cor  frontale  may  provide  direct 
information  on  blood  pressure  levels  which  could  be  used  by  either  the  supra- 
esophageal ganglion  or  the  tendon  ganglia. 

The  nerve  which  carries  the  two  axons  to  the  main  heart  valve  from  the 
supraesophageal  ganglion  appears  to  be  the  often  cited  nerve  of  Lemoine 
(Lemoine,  1868)  or  nervus  cardiacus  anterior  (Police,  1908;  Alexandrowicz, 
1932;  and  Health,  1941),  which  was  believed  by  them  to  originate  in  the  stomato- 
gastric ganglion.  However,  in  all  the  decapods  examined  in  the  present  study, 
these  two  axons,  stained  darkly  by  methylene  blue,  could  be  followed  clearly  from 
the  supraesophageal  ganglion  in  the  cor  frontale  nerve  down  the  length  of  the 
dorsal  median  artery  to  the  main  heart  valve  where  they  provide  the  sole  innerva- 
tion of  the  valve.  The  wall  of  the  dorsal  median  artery  along  its  entire  length 
is  a  meshwork  for  fine  nerve  fibers,  some  of  which  can  be  traced  to  the  stomato- 
gastric nerve  and  others  to  the  nerve  in  which  the  two  axons  run  to  the  heart 
valve.  In  some  cases,  distinct  connections  could  be  found  between  the  stomato- 
gastric nerve  and  the  nerve  from  the  cor  frontale  carrying  the  two  axons  to  the 
heart  valve.  These  connections  may  be  the  source  of  error  as  to  the  origin  of 
the  heart  valve  axons  in  the  earlier  literature. 

It  may  appear  strange  that  such  a  well  developed  system  as  the  cor  frontale 
has  previously  escaped  detailed  attention,  particularly  in  view  of  the  wide  interest 
in  crustacean  neurophysiology.  The  muscles  are  fairly  conspicious,  although 
they  had  been  previously  been  confused  with  the  eyestalk  muscles  judging  from 
their  inclusion  in  the  eyestalk  numbering  system  and  the  name,  musculi  oculi  basilis 
posterior,  applied  to  them  (Cochran,  1935).  However,  the  small  size  and  circuitous 
route  of  the  nerves  from  the  supraesophageal  ganglion  to  the  cor  frontale  and  the 
diffuseness  of  the  system  fin  comparison  to  the  simplicity  of  the  main  heart) 
may  also  explain  the  neglect.  In  addition,  electrophysiological  work  on  the 
supraesophageal  ganglion  in  an  isolated  preparation  has  been  hampered  by  the 
lack  of  proper  perfusion  techniques  fSteinacker,  1975)  and  so  (with  the  excep- 
tion of  recording  from  intact  animals)  the  cephalic  portion  of  the  decapods  has 
been  relatively  unexplored  in  comparison  to  the  extensive  work  on  more  pe- 
ripheral crustacean  ganglia. 


I  thank  Dr.  Donald  Kennedy  for  his  interest  and  encouragement,  Teppy 
Williams  for  illustrating  Figures  1  and  2  and  Jim  Brodal  for  Figure  3.  This 
work  was  supported  by  NIH  postdoctoral  fellowship  IFO  2  EY-55-012-01  and 
1F32EY-05-055-01. 

SUMMARY 

The  anatomy  of  an  auxiliary  heart  found  in  many  decapod  crustaceans  is 
described.  This  heart  is  found  at  the  anterior  end  of  the  dorsal  median  artery 


DECAPOD  CRUSTACEAN  AUXILIARY  HEART  507 

before  the  artery  divides  to  supply  the  cerebral  nervous  system.  The  heart  is 
essentially  two  strips  of  modified  somatic  muscle  located  inside  a  sinus  in  the 
dorsal  median  artery.  These  muscles  are  innervated  by  four  motoneurons  located 
in  the  supraesophageal  ganglion.  Sensory  innervation  and  possible  neurosecretory 
elements  are  also  described. 

LITERATURE  CITED 

ALEXANDROWICZ,  J.  S.,   1932.     The  innervation  of  the  heart  of  the  Crustacea.     I.     Decapoda. 

Q.  J.  Microsc.  Sd.,  75  :  181-249. 

BAUMANN,  H.  VON,  1917.     Das  cor  frontale  bei  decapoden  Krebsen.     Zool.  Anz..  49  :   137-144. 
COCHRAN,  D.  M.,  1935.     The  skeletal  musculature  of  the  blue  crab,  Callincctcs  sapidus  Rath- 
bun.     Snrithson.  Misc.  Pub!.,  92:   1-76. 
CROGHAN,  P.  C.,  1958.     The  survival  of  Artciuia  salina   (L. )   in  various  media.     J.  Exp.  Biol., 

35:  243-249. 
DANDO,  M.  R.,  AND  A.  I.  SELVERSTON,  1972.    Command  fibres  from  the  supraesophageal  ganglion 

to  the  stomatogastric  ganglion  in  Panulinis  argus.     J.   Couip.  Physio!.,  78:    138-175. 
DEBAISIEUX,  P.,  1944.     Les  yeux  de  crustaces.     Cellule,  50 :  9-122. 
DEMAL,  J.,  1953.     Genese  et  differenciation  d'hemocytes  chez  Palacnion  -curians  Teach.  Cellule, 

56:  85-102. 
FIELD,  L.  H.,  AND  J.  L.  LARIMER,  1975.     The  cardioregulatory  system  of  crayfish  :  the  role  of 

circumesophageal  interneurons.    /.  E.rp.  Biol.,  62:  531-543. 

Fox,  H.  M.,  1952.  Anal  and  oral  intake  of  water  by  Crustacea.  J.  E.vp.  Biol.,  29:  583-599. 
GOLDSTONE,  M.  W.,  AND  I.  M.  COOKE,  1971.  Histochemical  localization  of  monoamines  in  the 

crab  central  nervous  system.     Z.  Zcllforscli..   116:  7-19. 
HEATH,  J.  P.,  1941.     The  nervous  system  of  the  kelp  crab,  Pin/ettia  pradncta.     J.  Morphol., 

69:  481-498. 
LEMOINE,  V.,   1868.     Recherches  pour  servir  a  1'histoire  des  systemes  nerveux  musculaire  et 

glandulaire  de  1'Ecrevisse.     Ann.  Sci.  Xat.  Zool.,  9  :  99-280. 
MAYNARD,  D.  M.,  1960.     Circulation  and  heart  function.     Pages  161-226  in  T.  H.  Waterman, 

Ed.,  The  Physiology  of  Crustacea,  Vol.  /.     Academic  Press,  New  York  and  London. 
MENDELSON,  M.,  1971.     Oscillator  neurons  in  crustacean  ganglia.     Science,  171  :  1170-1173. 
PANTIN,  C.  F.  A.,   1969.     Notes  of  microscopical  technitiucs  for  zoologists.     Cambridge  Univ. 

Press,  London,  77  pp. 
POLICE,    G.,    1908.      Sul    sistema   nervosa   viscerale    dei    Crostacei   decapodi.      Mitt.    Zool.    Stat. 

Ncapel,   19:  69-116. 
SANDEMAN,    D.    C.,    1967.      The    vascular   circulation    in   the    brain,    optic    lobes    and    thoracic 

ganglia  of  the  crab,  Carcinus.  Proc.  R.  Soc.  Loud.  B.  Biol.  Sci..  168  :  82-90. 
STEINACKER,    A.,    1975.      Perfusion    of   the    central    nervous    system    of    decapod    crustaceans. 

Comp.  Biochcw.  Physiol.,  52A  :    103-104. 
STEINACKER,  A.,   1978.     Neural  and  neurosecretory  control  of  the  crustacean  auxiliary  heart. 

Am.  Zool.,  in  press. 
WEISMAN,    N.    A.,    1874.      Ueber    Bau   und   Lebenserscheinungen   von    Leptodora   hyalina.      Z. 

Wiss.  Zool.,  24:  349-418. 
WIERSMA,   C.  A.   G.,   AND  E.   NoviTSKi,   1942.     The   mechanism   of  the   nervous   regulation   of 

the  crayfish  heart.     /.  E.vp.  Biol.,  19:  255-265. 
WILKENS,   J.   L.,   L.   A.   WILKENS,   AND   B.   R.    McM.-\HON,    1974.     Central   control   of  cardiac 

and  scaphognathite  pacemakers  in  the  crab,   Cancer  iinniister.  J.   Comp.  Fh\siol.,  90: 

89-104. 


Reference:  Biol.  Bull.,  154:  508-516.     (June,  1978) 


RECTAL  GLAND  OF  FRESHWATER  STINGRAYS,  POTAMOTRYGON 
SPP.   (CHONDRICHTHYES :  POTAMOTRYGONIDAE) 

THOMAS    B.    THORSON,    ROBERT   M.    WOTTON.i   AND    TODD    A.    GEORGI 
School  of  Life  Sciences,   University  of  Nebraska — Lincoln,  Lincoln,  Nebraska  68588 

The  rectal  salt  gland  of  elasmobranchs  (also  known  in  the  English  literature 
as  the  caecal,  cloacal,  anal,  superanal,  rectal,  vermiform  and  digitiform  gland, 
process  or  appendage)  has  been  amply  treated  in  the  older  literature  (Hoskins, 
1917;  Crofts,  1925).  The  gland  has  been  likened  to  the  ink  sac  of  cephalopods, 
various  intestinal  diverticula  found  in  other  vertebrates  and  a  urinary  bladder, 
and  has  been  assigned  digestive,  reproductive,  secretory  and  blood-cleansing  func- 
tions. More  than  sixty  years  before  the  discovery  of  its  true  function,  Craw- 
ford (1899,  p.  60)  stated,  "The  rich  blood  supply,  the  character  of  the  secreting 
cells,  resembling  so  closely  as  they  do  the  cells  of  the  kidney,  and  the  occurrence 
of  urea  in  considerable  amount  in  the  secretion,  all  point  to  the  structure  having 
an  excretory  function,  and  playing  the  part  of  a  supplementary  kidney."  Craw- 
ford has  not  been  given  appropriate  credit  for  his  prescience,  which  was  essentially 
confirmed  by  Burger  and  Hess  (I960)  when  they  demonstrated  that  the  rectal 
gland  of  Sqnalus  acanthias  secretes  sodium  chloride  in  a  concentration  approxi- 
mately twice  that  of  the  plasma. 

Perhaps  because  of  the  technical  difficulties  in  collecting  the  rectal  gland 
fluid,  except  for  S.  acanthias,  it  has  been  collected  and  analyzed  only  from  the 
lip  shark,  HciniscyUium  plagioswn  (Chan,  Phillips  and  Chester  Jones,  1967)  and 
the  stingray,  Dasyatis  sabina  (Burger,  1972;  Beitz,  1977).  It  can  nevertheless 
be  reasonably  assumed  that  the  salt  secreting  function  of  the  rectal  gland  is 
universal  among  marine  elasmobranchs. 

Since  the  rectal  gland  functions  to  rid  the  body  of  excess  salt,  it  is  reason- 
able to  expect  that  secretion  would  stop  in  a  euryhaline  species  when  it  enters 
fresh  water.  Although  this  has  not  yet  been  conclusively  demonstrated,  Oguri 
(1964)  and  Gerzeli,  Gervaso  and  De  Stefano  (1969)  have  noted  that  the  rectal 
glands  of  Carcharhinus  leucas,  the  highly  euryhaline  bull  shark,  taken  from  a 
freshwater  environment,  are  smaller  than  in  the  same  species  taken  from  marine 
water.  Furthermore,  by  histological  examination,  they  noted  concomitant  regres- 
sive changes  in  the  secretory  tubules  of  the  freshwater  specimens. 

Carcharhinus  leucas  moves  back  and  forth  between  fresh  water  and  the  sea 
(Thorson,  1971)  and  can  readily  tolerate  both  media  (Thorson  and  Gerst,  1972). 
Presumably,  when  movement  takes  place  between  salt  and  fresh  water  the 
rectal  gland  alternates  between  activity  and  inactivity.  However,  the  family 
Potamotrygonidae,  freshwater  stingrays  of  South  American  river  systems,  live 
permanently  in  fresh  water  and  have  apparently  been  limited  to  fresh  water  for 
a  very  long  time.  They  no  longer  retain  high  concentrations  of  urea,  so  uni- 

1  Deceased  October  26,  1975. 

508 


RECTAL  GLAND  OF  FRESHWATER  STINGRAYS  509 

versally  employed  as  an  osmoregulatory  agent  by  marine  and  euryhaline  elasmo- 
branchs  (Thorson,  Cowan  and  Watson,  1967;  Junqueira,  Hoxter  and  Zago,  1968)  : 
nor  do  they  build  up  their  urea  content  when  transferred  to  varying  dilutions  of 
sea  water  (Thorson,  1970;  Griffith,  Pang.  Srivastava  and  Pickford,  1973;  Gerst 
and  Thorson,  1977).  The  urea  retaining  ability  is  apparently  of  no  further 
survival  value  to  elasmobranchs  in  a  freslrwater  environment.  Since  salts  are  in 
extremely  short  supply  in  the  fresh  water  of  tropical  South  American  rivers,  an 
interesting  question  is  posed  concerning  the  fate  of  the  strictly  freshwater  stingray's 
rectal  gland,  for  whose  salt-secreting  function  there  is  likewise  no  further  use. 

This  paper  presents  findings  concerning  the  morphological  aspects  of  this 
question,  as  well  as  some  of  their  physiological  connotations. 

MATERIALS  AND  METHODS 

Freshwater  stingrays  of  the  genus  Potauwtrygon  were  procured  at  Leticia, 
Colombia,  from  the  Amazon  River  drainage  in  Brazil  and  from  aquarium  sup- 
pliers in  the  United  States.  The  latter  specimens  were  imported  from  dealers  on 
the  Amazon  River  and  were  clearly  rays  of  the  subject  genus,  although  identifica- 
tion to  species  was  not  always  possible. 

Sections  were  made  of  portions  of  appropriate  tissues  from  numerous  rays, 
but  the  illustrations  and  discussion  are  based  primarily  on  two :  specimen  A,  a 
juvenile  female  (Potamotrygon  uiotoro)  of  160  mm  disc  width,  purchased  from 
a  Nebraska  supplier ;  and  specimen  B,  a  female  approaching  sexual  maturity  (P. 
circularis),  413  mm  disc  width,  taken  by  a  local  dealer  from  the  Itacoai  River, 
an  Amazon  tributary  in  extreme  western  Brazil,  near  Leticia,  Colombia. 

From  the  rays  selected  for  study,  a  section  of  the  lo\ver  end  of  the  alimentary 
canal,  with  associated  structures,  was  immersed  either  directly  in  Bouin's  fluid 
or  in  lO^o  formalin  and  later  transferred  to  Bouin's.  Specimens  were  trans- 
ferred through  several  changes  of  70/o  alcohol  to  remove  the  excess  Picric  acid. 
All  the  tissues  were  passed  through  successive  increasing  strengths  of  alcohol  up 
to  absolute,  to  insure  thorough  dehydration.  The  last  change  of  1009o  alcohol 
was  replaced  with  xylene  to  which  dryrite  had  been  added.  After  clearing,  they 
were  placed  in  molten  paraffin  and  subsequently  embedded  in  wax.  Sections  \vere 
cut  at  8  to  10  micra  and  the  ribbons  affixed  to  slides  with  Meyer's  albumin 
fluid.  The  sections  were  stained  writh  Ehrlich's  acid  hematoxylin  and  Eosin  as 
tinctorial  agents,  and  mounted  in  balsam  under  glass  cover  slips. 

Micrographs  were  taken  with  a  Zeiss  photo-microscope  II  on  Ilford  Pan  F 
film. 

RESULTS 

The  rectal  gland  and  associated  tissues  of  specimen  A  (a  juvenile  female 
Potainotrygon  motoro)  are  shown  in  Figure  1.  The  gland  is  a  short,  slender 
structure,  directed  anteriorly  from  the  dorsal  side  of  the  post-valvular  intestine. 
It  is  closely  associated  with  three  ovoid  masses  of  white  tissue.  Both  the  gland 
and  the  three  white  lobes  are  covered  with  peritoneum.  The  same  structures  and 
arrangement  were  found  in  specimen  B,  a  female  P.  circitlaris  approaching  sexual 


510 


THORSON,  WOTTON  AND  GEORGI 


FIGURE  1.  Rectal  gland  and  associated  myeloid  lobes  of  a  juvenile  Potamotrygon  motor o 
(160  mm  disc  width).  The  gland  is  the  slender  structure  at  left. 

maturity  (Fig.  2).  In  this  larger  animal  the  white  masses  have  become  further 
lobed  and  irregular  in  shape. 

A  representative  cross  section  of  the  rectal  gland  from  specimen  A  is  shown 
in  Figure  3  and  an  enlarged  area  of  the  section  in  Figure  4.  The  glandular 
portion,  surrounding  a  central  lumen  (A),  occupies  approximately  the  central 
half  of  the  gland's  diameter.  It  includes  a  series  of  tubules  (B)  which  are 
composed  of  simple  cuboidal  cells  and  drain  into  the  central  lumen,  as  at  C. 
Surrounding  the  central  glandular  portion  and  forming  most  of  the  remainder  of 
the  gland,  is  a  broad  band  of  connective  tissue  containing  blood  vessels  (D) 
and  sinuses.  The  free  surface  of  the  gland  is  covered  by  a  stratified  columnar 
epithelium  (E).  The  central  lumen  is  lined  with  a  simple  squamous  epithelium 
which  becomes  stratified  as  it  comes  closer  to  and  enters  the  rectum. 

The  gland,  although  in  close  connection  with  the  associated  lobes  (Fig.  1), 
is  clearly  and  completely  independent,  being  separated  from  them  by  a  broad 
layer  of  connective  tissue  (Fig.  3). 

Examination  of  the  lobes  associated  with  the  rectal  gland  discloses  an  external 
epithelial  layer  continuous  with  that  of  the  gland.  The  epithelium  covers  a  thin 
connective  tissue  stratum  and  inside  is  a  heavy  concentration  of  leucocytes  of 
various  kinds  and  stages,  including  some  with  mitotic  figures. 

DISCUSSION 

The  white  lobes  associated  with  the  rectal  gland  undoubtedly  represent  the 
"lymphoid  tissue"  described  and  figured  in  several  earlier  accounts  of  the  gross 
structure  of  the  rectal  gland  (e.g.,  Hoskins,  1917).  Their  histology  indicates 


RECTAL  GLAND  OF  FRESHWATER  STINGRAYS 


511 


FIGURE  2.     Rectal  gland    (rg)    with  portion  of  rectum    (r)    and  associated  myeloid  lobes 
(m)   of  a  female  Potainotrygon  circularis  nearing  sexual  maturity   (413  mm  disc  width). 

that  they  are  a  part  of  the  lymphomyeloid  system  of  cartilaginous  fishes  recently 
discussed  by  Fange  (1977).  This  system  is  active  in  haemopoiesis  and  in  the 
immune  responses.  Components  of  the  system  mentioned  by  Fange  include  the 
spleen  and  thymus;  the  epigonal  organs  (associated  with  the  gonads)  ;  Leydig's 
organ  in  the  esophagus;  extensive  tissue  in  the  cranium  (in  holocephalans)  ;  and 
aggregations  of  leucocytes  in  the  connective  tissue  of  the  kidneys  and  the  intestine 
(spiral  valve). 

The  prominence  and  distinctness  of  the  organs  discussed  here  and  their  close 


FIGURE  3.     Cross  section  of  rectal  gland  shown  in  Figure  1. 


51_>  THORSON,  WOTTON  AND  GEORGI 


O.lmm 


A 
C 


. 


B 


FIGURE  4.  Enlarged  area  of  rectal  gland  section  (from  Fig.  3)  :  a  central  lumen  (A), 
lined  with  simple,  squamous  epithelium,  is  surrounded  by  scattered  tubules  (B)  which  empty 
into  lumen,  as  at  C.  A  wide  band  of  connective  tissue,  with  blood  vessels  (D)  occupies  the 
outer  portion  of  the  section,  and  the  gland  is  covered  with  a  stratified,  columnar  epithelium  (E). 

association  with  the  rectal  gland  and  post-valvular  gut  justify  their  designation 
as  rectomyeloid  bodies. 

Goldstein  and  Forster  (1971)  were  unable  to  find  a  rectal  gland  in  Potamo- 
trygon sp.  Griffith  et  al.  (1973)  reported  that  rays  studied  by  them  (Potamotrygon 
spp. )  had  an  organ  in  the  anatomical  position  of  the  rectal  gland,  but  histological 
investigation  showed  that  it  was  structurally  unlike  the  rectal  gland  of  marine 
elasmobranchs.  Gerzeli  et  al.  (1969)  and  Gerzeli,  De  Stefano,  Bolognani,  Koenig, 
Gervaso  and  Omodeo-Sale  (1976)  reported  a  rectal  gland  in  a  stingray  identified 
as  Potamotrygon  brucJiyun/s  and  noted  that  it  was  large,  with  a  gland  weight/ 
body  weight  ratio  of  ca.  1  X  10~3.  Gerst  and  Thorson  (1977)  reported  the 
presence  of  a  structure  in  Potamotrygon  spp.  with  the  location  and  histological 
characteristics  of  the  elasmobranch  rectal  gland,  but  of  reduced  proportions. 

The  present  study  conclusively  establishes  the  presence  of  a  rectal  gland  in 
the  Potamotrygonidae,  with  the  location  and  histological  features  of  the  gland 
found  in  marine  elasmobranchs.  The  confusion  and  conflicting  reports  cited  above 
may  be  explained  by  the  fact  that  the  gland  is  small  and  inconspicuous  and  may 
easily  be  obscured  by,  or  even  mistaken  for,  a  part  of  the  lobed  myeloid  tissue 
spatially  associated  with  it. 

Marine  and  brackishwater  elasmobranchs  have  relatively  large  glands  (Table  I). 
The  fully  euryhaline  shark  (Carcharliinns  Iciicas)  that  completely  tolerates  both 


RECTAL  GLAND  OF  FRESHWATER  STINGRAYS  513 

TABLE  I 

Rectal  aland/body  weight  ratio  related  to  habitat. 

Rectal  gland 
\vt.   body  wt.  ratio 
Reference  Species  (units  per  million) 

Marine  and  brackish  water  species 

Burger  (1972)  Squalus  acanthias  600 

Dasyatis  sabina    j  240 

Fange  and  Fugelli  (1963)  Selache  maxima  -!<><i 

Chan  and  Phillips  (1967)  Hemiscyllium  plagiosum  120 

Renting  (1966)  Squalus  acanthias  444 

Carcharius  littoral  is  185 

Carcharhinus  falciform  is  1 90 

Mustelus  canis  214 

Raja  eglanteria  202 

Myliobatiis  freiiiinvillei  164 

Pteroplatea  altavela  86 

Squatina  squatina  99 

Fully  euryhaline  species 
Gerzeli  el  al.  (1969)  Carcharhinus  leucas  60 

(marine) 
Carcharhinus  leucas  20 

(fresh  water) 

Thorson  (unpublished)  Carcharhinus  leucas  30 

(fresh  water) 

Freshwater  species 

Thorson  (unpublished)  Potamotrygon  circularis  15 

Gerzeli  et  al.  (1969)  Potamotrygon  brachyurus  1000* 

*  See  text. 

fresh  and  salt  water  has  a  gland  of  somewhat  reduced  relative  size ;  the  gland 
appears  to  be  larger  when  this  shark  is  in  sea  water  than  when  it  is  in  fresh 
water  (Oguri,  1964;  Gerzeli  et  al.,  1969).  The  completely  freshwater  rays 
(Potamotrygonidae)  examined  in  this  study  have  rectal  glands  of  still  more 
reduced  size.  Furthermore,  the  number  of  tubules  is  considerably  reduced  and 
their  distribution  within  the  gland  is  relatively  restricted  (Fig.  3).  The  rectal 
gland  weight/body  weight  ratio  of  I  X  1O3  given  by  Gerzeli  et  al.  (1969)  is  greater 
than  that  of  any  marine  elasmobranch  listed  in  Table  I.  The  figure  must  either  be 
in  error  or  the  specimen  studied  may  have  included  myeloid  or  other  tissue  in 
addition  to  the  gland  itself. 

Atrophy  of  the  gland  might  reasonably  be  expected  in  rays  that  have  been 
completely  limited  to  fresh  water  for  a  long,  although  undetermined,  period  of 
time.  Just  as  they  have  abandoned  urea  retention,  the  freshwater  rays  apparently 
have  also  abandoned  supplementary  salt  excretion.  Both  would  be  counter- 
productive in  a  freshwater  environment. 

The  highly  euryhaline  Carcharhinus  leucas  is  able  to  increase  and  decrease  the 
urea  content  of  its  body  fluids  in  response  to  changes  in  environmental  salinity. 
The  findings  of  Oguri  (1964).  and  Gerzeli  et  al.  (1969;  1976)  suggest  that  secre- 
tory activity  of  the  rectal  gland  of  C.  leucas  also  responds  to  changes  in  environ- 


514  THORSON,  WOTTON  AND  GEORGI 

mental  salinity.  In  Potamotrygon  spp.,  on  the  other  hand,  transfer  to  saline 
environment  does  not  elicit  an  increase  in  urea  concentration  in  body  fluids,  and 
the  loss  of  urea  retention  appears  to  be  irreversible  (Thorson,  1970;  Griffith 
i1/  ill,  1('73;  Gerst  and  Thorson.  1977).  This  fact  suggests  that  the  apparent 
loss  of  salt  secretory  activity  of  the  atrophied  rectal  gland  in  Potamotrygon  may 
also  be  irreversible.  This  view  is  supported  by  the  observation  that,  in  potamotry- 
gonids  transferred  to  dilute  sea  water,  regulation  of  inorganic  ions  breaks  down. 
Sodium  and  chloride  concentrations  in  particular  almost  double  in  seawater- 
acclimated  rays.  The  greatest  concentration  they  can  tolerate  for  any  length 
of  time  is  approximately  40%  sea  water  (Thorson  ct  al,  1967  \  Thorson,  1970; 
Griffith  ct  al.,  1973;  Gerst  and  Thorson,  1977). 

No  function,  other  than  salt  secretion,  has  been  demonstrated  for  the  fully 
active  rectal  gland  of  marine  elasmobranchs.  What  residual  function  the  atrophied 
potamotrygonid  rectal  gland  might  have,  if  any,  is  unknown.  Gerzeli  et  al., 
(1976,  p.  619)  reported  that  the  rectal  gland  of  Potamotrygon  brachyurus  "appears 
very  peculiar,  showing  secretory  activity  histologically,  but  lacking  any  cytochemical 
evidence  related  to  salt  secretion."  Otherwise,  nothing  has  been  established 
experimentally  or  histochemically  concerning  any  specific  function  for  the 
potamotrygonid  rectal  gland. 

The  Chondrichthyes  made  their  appearance  in  the  geological  record  during 
the  Devonian.  Although  their  presumed  ancestors,  the  placoderms,  may  have 
inhabited  inland  fresh  waters,  the  Devonian  Chondrichthyes  appear  to  have  been 
marine  since  their  first  appearance  (Romer,  1966).  This  study  does  little  to 
elucidate  the  continuing  discussion  of  whether  urea  retention  developed  in 
chondrichthians  in  response  to  the  invasion  of  salt  water  or  existed  earlier  and 
provided  a  pre-aclaptation  for  marine  life.  However,  it  provides  evidence  bearing 
on  a  related  question  concerning  the  potamotrygonid  stingrays :  does  the  near 
absence  of  urea  in  the  freshwater  rays  represent  a  genetic  deletion  of  their  ancestral 
ability  to  concentrate  urea,  or  were  they  descended  from  ancestors  that  had  never 
left  fresh  water  and  had  never  developed  urea  retention?  The  latter  possibility 
was  considered  remote  by  Thorson  ct  al.  (1967)  and  less  plausible  than  the  former 
by  Forster  and  Goldstein  (1969).  However,  evidence  concerning  urea  retention 
is  not  preserved  in  fossils,  and  there  is  in  any  case  no  fossil  record  of  the  family 
Potamotrygonidae.  Stingrays  of  the  closely  related  marine  family  Dasyatidae  are 
known  from  freshwater  assemblages  of  the  Tertiary  (Eocene),  but  reports  of 
fossil  Potamotrygonidae  in  South  America  (Carman,  1913)  and  in  Africa  (Aram- 
bourg,  1947)  are  probably  also  of  dasyatids  (Thorson  and  Watson,  1975). 

In  the  absence  of  fossil  evidence  of  the  history  of  this  group,  evidence  must 
be  sought  from  extant  rays.  Such  evidence  is  now  provided  by  the  rectal  gland, 
which  is  likewise  not  preserved  in  fossils.  A  functional  rectal  gland  can  only  be 
viewed  as  a  marine  adaptation  and  its  presence,  albeit  in  much  reduced  form,  with 
no  known  function,  can  only  indicate  a  marine  ancestry  for  the  freshwater 
stingrays.  The  chronology  of  the  gland's  earliest  history  cannot  at  present  be  firmly 
established,  but  at  the  time  of  the  first  appearance  of  the  stingrays,  in  the  Cretace- 
ous (Romer,  1966),  they  were  almost  certainly  already  marine,  as  were  the  other 
Chondrichthyes,  and  possessed  the  functional  rectal  gland  so  universally  found  in 
the  other  cartilaginous  fishes. 


RECTAL  GLAND  OF  FRESHWATER  STINGRAYS  515 

Both  the  absence  of  urea  retention  and  the  atrophy  of  the  rectal  gland  bespeak 
a  long  history  in  fresh  water  for  the  Potamotrygonidae.  The  salinity  tolerance, 
urea  retaining  ability  and  the  size  and  condition  of  the  rectal  gland,  studied  in  a 
variety  of  stingray  species  representing  the  full  spectrum  of  environmental  salinities, 
are  potentially  rich  sources  of  evidence  regarding  the  evolution  of  freshwater 
adaptation  in  stingrays  as  well  as  elasinobranchs  in  general. 

The  study  was  supported  in  part  by  NIH  Grant  HE-09075,  the  University 
Research  Council  of  the  University  of  Nebraska-Lincoln  and  the  National  Geo- 
graphic Society.  The  photograph  in  Figure  3  was  made  by  Harley  Ridgeway. 

SUMMARY 

1.  Contrary  to  some  reports,  a  rectal  gland  is  present  in   strictly  freshwater 
stingrays  of  South  American  rivers  (Potamotrygon  spp.). 

2.  The  gland  has  the  location   and  histological   features   of  the   salt-secreting 
rectal  gland  of  marine  elasinobranchs,  but  is  much  reduced  in  size  and  number 
of  tubules. 

3.  Its  residual  function,  if  any,  is  unknown. 

4.  The  rectal  gland  is  associated  writh  prominent  myeloid  lobes,  here  designated 
as  rectomyeloid  bodies. 

5.  In  the  absence  of  potamotrygonid  fossils,  the  atrophied  rectal  gland  is  strong 
evidence  of  marine  ancestry  for  the  freshwater  rays. 

6.  Both  the  reduced  gland  and  the  loss  of  urea  retention  in  potamotrygonids 
are  indicative  of  a  long  history  of  freshwater  adaptation. 

LITERATURE  CITED 

ARAMBOUKG,  C,  1947.  Mission  scientifique  de  1'Omo  (1932-1933),  1:  469-471.  (Museum 
National  d'Historie  Naturelle,  Paris). 

BEITZ,  B.  E.,  1977.  Secretion  of  rectal  gland  fluid  in  the  Atlantic  stingray,  Dasyatis  sabina. 
Copcia,  1977(3):  585-587. 

BONTING,  S.  L.,  1966.  Studies  on  sodium-potassium-activated  adenosinetriphosphatase.  XV. 
The  rectal  gland  of  the  elasmobranchs.  Comp.  Biochem.  Physiol.  17  :  953-966. 

BURGER,  J.  W.,  1972.  Rectal  gland  secretion  in  the  stingray,  Dasyatis  sabina.  Comp.  Biochem. 
Physiol,  42 A:  31-32. 

BURGER,  J.  W.,  AND  W.  N.  HESS,  1960.  Function  of  the  rectal  gland  in  the  spiny  dogfish. 
Science,  131  :  670-671. 

CHAN,  D.  K.  O.,  AND  J.  G.  PHILLIPS,  1967.  The  anatomy,  histology  and  histochemistry  of 
the  rectal  gland  in  the  lip-shark  Hemiscyllvuvn  plagiosiiiii  (Bennett).  /.  Anat.,  101: 
137-157. 

CHAN,  D.  K.  O.,  J.  G.  PHILLIPS,  AND  I.  CHESTER  JONES,  1967.  Studies  on  electrolyte  changes 
in  the  lip-shark,  Hemiscyllium  plagiosnm  (Bennett),  with  special  reference  to  hor- 
monal influence  on  the  rectal  gland.  Comp.  Biochem.  Physiol..  23:  185-198. 

CRAWFORD,  M.  B.,  1899.  On  the  rectal  gland  of  the  elasmobranchs.  Proc.  R.  Soc.  Edinburgh, 
23:  55-61. 

CROFTS,  D.  R.,  1925.  The  comparative  morphology  of  the  caecal  gland  (rectal  gland)  of 
selachian  fishes,  with  some  reference  to  the  morphology  and  physiology  of  the  similar 
intestinal  appendage  throughout  Ichthyopsida  and  Sauropsida.  Proc.  Zool.  Soc.  Lon- 
don, 1925:  101-188. 

FANGE,  R.,  1977.  Size  relations  of  lymphomyeloid  organs  in  some  cartilaginous  fish.  Actti 
Zool.,  58:  125-128. 


516  THORSON,  WOTTON  AND  GEORGI 

FANGE,  R.,  AND  K.  FUGELLI,  1963.    The  rectal  salt  gland  of  elasmobranchs,  and  osmoregulation 

in  chimaeroid  fishes.    Sarsiti,  10  :  27-34. 
FORSTER,  R.  P.,  AND  L.  GoLDSTKiN,  1969.     Formation  of  excretory  products.     Pages  313-350  in 

W.  S.  Hoar  and  D.  J.  Randall,  Eds.,  Fish  physiology,  Vol.  I  Excretion,  ionic  regula- 
tion and  metabolism.     Academic  Press,  New  York. 

CARMAN,  S.  W.,  1913.    The  Plagiostomia.    Mem.  Mus.  Com  p.  Zool.  Harvard,  36:  1-515. 
GERST,  J.  W.,  AND  T.  B.  THORSON,  1977.     Effects  of  saline  acclimation  on  plasma  electrolytes, 

urea  excretion,  and  hepatic  urea  biosynthesis  in  a  freshwater  stingray,  Potamotrygon  sp. 

Carman,  1877.     Comp.  Biochcm.  Physiol.,  56A :  87-93. 
GERZELI,  G.,  M.  V.  GERVASO,  AND  G.  F.  DE  STEFANO,  1969.     Aspetti  della  ghiandola  rettale  e 

della  regolazione  osmotica  in  Selaci  marini  e  d'acqua  dolce.     Bol.  Zool.,  36  :  399-400. 
GERZELI,  G.,  G.  F.  DE  STEFANO,  L.  BOLOGNANI,  K.  W.  KOENIG,  M.  V.  GERVASO,  AND  M.  F. 

OMODEO-SALE,   1976.     The  rectal  gland  in  relation  to  the  osmoregulatory  mechanisms 

of   marine   and   freshwater   elasmobranchs.      Pages   619-627    in    T.    B.    Thorson,    Ed., 

Investigations  of   the   ichthyofauna    of  Nicaraguan   lakes.      School    of    Life    Sciences, 

University  of   Nebraska-Lincoln. 
GOLDSTEIN,  L.,  AND  R.  P.  FORSTER,   1971.     Urea  biosynthesis  and  excretion  in  freshwater  and 

marine  elasmobranchs.     Coin  p.  Biochem.  Physiol.,  39B  :  415-421. 
GRIFFITH,  R.  W.,   P.   K.  T.   PANG,  A.   K.   SRIVASTAVA,  AND  G.   E.   PICKFORD,    1973.     Serum 

composition  of  freshwater  stingrays    (Potamotrygonidae)    adapted  to  fresh  and  dilute 

sea  water.     Biol.  Bull.,   144:  304-320. 
HOSKINS,   E.   R.,    1917.      On  the   development   of  the   digitiform   gland   and   the   post-valvular 

segment  of  the  intestine  in  Squahis  acanthias.     J.  Morphol.,  28  :   329-367. 
JUNQUEIRA,  L.  C.  U.,   G.  HOXTER,  AND  D.  ZAGO,   1968.     Observations  on  the  biochemistry  of 

fresh  water  rays  and  dolphin  blood  serum.    Rev.  Bras.  Pcsquis.  Med.  Biol.,  1  :  225-226. 
OGURI,   M.,    1964.     Rectal   glands   of  marine   and  freshwater    sharks :    comparative   histology. 

Science,  144:   1151-1152. 
PANG,  P.  K.  T.,  R.  W.  GRIFFITH,  AXD  J.  W.  ATZ,   1977.     Osmoregulation  in  elasmobranchs. 

Am.  Zool.,  17:  365-377. 
ROMER,   A.    S.,    1966.      Vertebrate   paleontology.    University   of    Chicago    Press,    Chicago    and 

London,  468  p. 
THORSON,  T.  B.,  1970.     Freshwater  stingrays,  Potamotrygon  spp. :  failure  to  concentrate  urea 

when  exposed  to  saline  medium.     Life  Sci.,  9  :  893-900. 
THORSON,  T.  B.,  1971.    Movement  of  bull  sharks,  Carcharhinus  leucas,  between  Caribbean  Sea 

and   Lake   Nicaragua  demonstrated   by  tagging.     Copcia,   1971(2)  :   336-338. 
THORSON,   T.   B.,  AND  J.   W.   GERST,    1972.     Comparison  of  some   parameters  of   serum   and 

uterine    fluid   of   pregnant,    viviparous    sharks    (Carcharhinus    leucas*)    and    serum    of 

their  near-term  young.     Coin  p.   Biochcm.  Physiol.,  42A :   33-40. 
THORSON,  T.  B.,  AND  D.  E.  WATSON,  1975.     Reassignment  of  the  African  freshwater  stingray, 

Potamotrygon  garouaensis,   to   the   genus   Dasyatis,   on   physiologic   and   morphologic 

grounds.     Copeia,  1975(4)  :  701-712. 
THORSON,  T.  B.,  C.  M.  COWAN,  AND  D.  E.  WATSON,  1967.    Potamotrygon  spp.  :  elasmobranchs 

with  low  urea  content.     Science,   158  :   375-377. 


INDEX 


Acclimation,  in  a  copepod,  177 

ACHE,  B.  W.  See  Z.  M.  Fuzessery,  226 

Acid  phosphatase,  in  life  cycle  of  Panagrellus 

silusiae,  374 

Aging,  and  the  relative  activity  of  acid  phos- 
phatase isozymes  in  a  nematode,  374 
ALKON,  D.  L.  See  J.  F.  Harrigan,  430 
Amino  acid  uptake,  in  Dendraster  excentricus, 

335 
Ammonia,    effect    of    ionized    and    un-ionized, 

on  growth  of  prawn  larvae,  15 
toxicity  to  larval  shrimp,  15 
Amphibian  reproduction,  198 
Amphioplus  abditus,  skeletal  development,   79 
ANDERSON,  J.   M.  Studies  on  functional  mor- 
phology in  the  digestive  system  of  Oreaster 
reticulatus  (L.)  (Asteroidea),  1 
ANDERSON,  S.  See  D.  E.  Morse,  440 
Annual    reproductive    cycle,    of    Leptosynapta 

tenuis,  68 
Antennular  chemosensitivity,  in  spiny  lobster, 

226 

Aplidium  multiplicatum,  budding  behavior,  453 
Apodous  holothurian,  gonad  development  in,  68 
Aporrhais,  burrowing  behavior,  463 
ARMSTRONG,  D.  A.,  D.  CHIPPENDALE,  A.  W. 
KNIGHT,  AND  J.  E.  COLT.   Interaction  of 
ionized  and  un-ionized  ammonia  on  short- 
term  survival  and  growth  of  prawn  larvae, 
Macrobrachium  rosenbergii,  15 
Ascidians,  budding  behavior,  453 
Asteroidea,  digestive  system,  1 
Auxiliary  heart,  crustacean,  anatomy,  497 
Axenic  culture  of  Cladocera,  47 

B 

BACKLUND,  P.  S.  See  G.  C.  Stephens,  335 
Balanus  glandula,  reproduction,  262 
BARKER,  M.  F.  Descriptions  of  the  larvae  of 
Slichaster    austral  is    (Verrill)    and    Cosci- 
nasterias    calamaria     (Gray)     (Echinoder- 
mata:    Asteroidea)    from    New    Zealand, 
obtained  from  laboratory  culture,  32 
Barnacles,  reproduction,  262 
Barometric    pressure,    influence   on    locomotor 

activity  levels  in  Rana  pipiens,  302 
Behavior,   adult   egg-laying,   of   Nassarius   ob- 

soletus,  282 

budding,  in  ascidians,  453 
burrowing,  of  Aporrhais,  463 


descriptions  of  feeding,  in  Oreaster  reticula- 
tus, 1 

Bioenergetics,  of  Littorina  irrorata,  322 
Biology,  of  Carcinonemertes  epialti,  121 
Biomphalaria  glabrata,  chemoreception  and 

rheotaxis  in,  361 

Bivalve  mantle,  water  permeability,  292 
Bottlenose  porpoise  (Tur stops  truncatus)  group 

organization,  348 

BOUSFIELD,  J.   D.   kheotaxis  and  chemorecep- 
tion in  the  freshwater  snail  Biomphalaria 
glabrata   (Say) :  estimation  of  the  molecu- 
lar weights  of  active  factors,  361 
BRADLEY,  B.  P.  Increase  in  range  of  tempera- 
ture    tolerance     by     acclimation     in     the 
copepod  Eurytemora  affinis,  177 
Brood  production,  in  intertidal  barnacles,  262 
Budding  experiments,  on  ascidians,  453 
Burrowing  behavior,  of  Aporrhais,  463 


Callianassa   sp.,   osmotic  and    ionic  regulation 

in,  409 

Calving  seasonally,  porpoise,  348 
Carcinonemertes  epialti,  biology,  121 
Cardiac  stomach,  in  Oreaster  reticulatus,  1 
CARR,  W.  E.  S.  See  Z.  M.  Fuzessery,  226 
Chemoreception,  in  freshwater  snails,  molecu- 
lar  weight   characteristics  of  attractants, 
361 

in  spiny  lobster,  226 
CHILDRESS,  J.  J.  See  T.  J.  Mickel,  138 
CHIPPENDALE,  D.  See  D.  A.  Armstrong,  15 
Chondrichthyes,  rectal  gland,  508 
Chromatophorotropic  activity,  of  CNS  extract 

from  Limulus,  148 
Chthamalus  fissus,  reproduction,  262 
Circadian  rhythm,   in  Littorina  irrorata  respi- 
ration, 322 

Cladocera,  culture  in  artifical  media,  47 
Coelenterates,  prostaglandin  synthetase,  440 
COLT,  J.  E.  See  D.  A.  Armstrong,  15 
CONKLIN,   D.  E.  AND  L.  PROVASOLi.  Biphasic 
particulate  media  for  the  culture  of  filter- 
feeders,  47 
Constant  acclimation   temperature,   effects  of, 

in  crabs,  188 

Copepod,  acclimation  in,  177 
Corals,  prostaglandin  synthetase  in,  440 
Corymorpha  bigelowi,  life  cycle,  485 
Coscinasterias  calamaria,  descriptions  of  larvae 
of,  32 


517 


518 


1NDKA 


Crustacea:    Decapoda,    nioutli parts   and    setae 

of  larval  lobsters,  383 

Crustacea,  effect  of   pH   on   oxygen  consump- 
tion, 138 
Crustacean  auxiliary  heart,  anatomy,  497 

cardiac  system,  anatomy,  497 

muscle,  development  in  juvenile  lobsters,  55 
Crustaceans,  ammonia  toxicity  in  culture  and 

maintenance  of,  15 
Crab  respiration,  188 

Cultivation,  of  Hermissenda  crassicornis,  430 
Culture,  of  filter-feeders  in  artificial  media,  47 
Cuthona  nana,  development  and  ecology  of,  157 
Cyclic  acclimation   temperature,   effects  of   in 
crabs,  188 

D 

DAME,  R.  F.  AND  F.  J.  VERNBERG.  The  in- 
fluence of  constant  and  cyclic  acclimation 
temperatures  on  the  metabolic  rates  of 
Panopeus  herbstii  and  Uca  pugilator,  188 

Decapod  crustacean,  larval  development,  241 

DEL  PINO,  E.  M.  AND  A.  A.  HUMPHRIES,  JR. 
Multiple  nuclei  during  early  oogenesis  in 
Flectonotus  pygmaeus  and  other  marsupial 
frogs,  198 

Dendraster  excentricus,  uptake  of  amino  acids, 
335 

DENOUX,  G.  J.  See  T.  C.  Shirley,  322 

Desiccation,  during  development  of  Nassarinx 
obsoletus,  282 

Development,  of  larvae  of  Stichaster  austral  is 

and  Coscinasterias  calamaria,  32 
of  the  mouthparts  of  larval  lobsters,  383 
of  the  nudibranch  Cuthona  nana,  157 

Diet,  of  Hermissenda  crassicornis,  430 

Digestive  system,  of  Oreaster,  functional  mor- 
phology, 1 

Dive  times,  porpoise,  348 

DOERING,  G.  N.  AND  E.  E.  PALINSCAR.  Acid 
phosphatase  during  the  life  cycle  of  the 
nematode,  Panagrelhis  silusiae,  374 

DORES,  R.  M.  See  P.  D.  Pezalla,  148 

E 

Echinoderm,  development  of  Asteriodea  larvae, 

32 

gonad  development  in,  68 
skeletal  ontogeny  and  phylogeny,  79 
Ecology,  of  the  nudibranch  Cuthona  nana,  157 
Egg  capsules,  Nassarius  obsoletus,  282 
Egg  predation,  on  ovigerous  crabs,  by  Carct- 

nonemertes,  121 

Elasmobranch  osmoregulation,  508 
Electron  microscopy,  of  Linens  ruber,  213 
Electrophysiology,   of  taurine  sensitive  recep- 
tors in  spiny  lobster  antennules,  226 


Epidermal  absorption,  by  the  rhynchocoelan, 

Linens  ruber,  213 

Epifaunal  activity,  of  Aporrhais,  463 
Euphysora,  life  cycle,  485 
Eurylemora  affinis,  acclimation  in,  177 


FACTOR,  J.  R.  Morphology  of  the  mouthparts 
of    larval    lobsters,    Homarus    americanus 
(Decapoda:    Nephropidae),    with    special 
emphasis  on  their  setae,  383 
Fecundity,  of  Hermissenda  crassicornis,  430 
Feeding,  in  Phoronopsis  viridis,  472 
FELDER,   D.   L.  Osmotic  and  ionic  regulation 
in  several  Western  Atlantic  Callianassidae 
(Crustacea,  Decapoda,  Thalassinidea),  409 
Ficopomatus,  generic  revision  of,  96 
Filter-feeders,    particulate    media    for    culture 

of,  47 

FISHER,  F.  M.,  JR.  AND  J.  A.  OAKS.  Evidence 
for  a  nonintestinal  nutritional  mechanism 
in  the  rhynchocoelan,  Linens  ruber,  213 
Flectonotus  pygmaeus,  multinucleate  oogenesis, 

198 

Food-resources,  of  Phoronopsis  viridis,  472 
Freshwater  stingrays,  rectal  gland,  508 
Frogs,  marsupial,  multinucleate  oogenesis,  198 
FUZESSERY,  Z.  M.,  W.  E.  S.  CARR,  and  B.  W. 
ACHE.  Antennular  chemosensitivity  in  the 
spiny  lobster,  Panulirus  argus:  studies  of 
taurine  sensitive  receptors,  226 


Gametogenesis,  in  Holothuroidea,  68 
Gastropod   burrowing  behavior,   of  Aporrhais, 

463 

Generic  revision,  of  Ficopomatus,  96 
Geographic  range,   of   Carcinonemertes  epialti, 

121 

GEORGI,  T.  A.  See  T.  B.  Thorson,  508 
Gnathophausia  ingens,  effect  of  pH  on  oxygen 

consumption,  138 

Gorgonians,  prostaglandin  synthetase,  440 
GOVIND,   C.    K.   AND   F.    LANG.    Development 

of   the  dimorphic  claw  closer   muscles  of 

the     lobster    Homarus    americanus.     III. 

Transformation   to   dimorphic   muscles   in 

juveniles,  55 
GOY,  J.  W.  AND  A.  J.  PROVENZANO,  JR.  Larval 

development  of  the  rare  burrowing  mud 

shrimp   Naushonia  crangonoides   Kingsley 

(Decapoda:  Thalassinidea;  Laomediidae), 

241 
GREEN,  J.   D.  The  annual  reproductive  cycle 

of  an   apodous   holothurian,   Leptosynapta 

tennis:  a  bimodal  breeding  season,  68 
Group  composition,  porpoise,  348 
Growth  inhibition  in  larval  shrimp,  15 


INDEX 


519 


H 


HALL,  C.  See  R.  D.  Prusch,  292 

HARRIGAN,  J.  F.  AND  D.  L.  ALKON.  Larval 
rearing,  metamorphosis,  growth  and  re- 
production of  the  eolid  nudibranch  Her- 
missenda crass icorn is  (Eschscholtz,  1831) 
(Gastropoda:  Opisthobranchia),  430 

Hemigrapsus  oregonensis,  host  for  Carcinone- 
mertes,  121 

HENDLER,  G.  Development  of  Amphioplus  ab- 
ditus  (Verrill)  (Echinodermata:  Ophiu- 
roidea).  II.  Description  and  discussion  of 
ophiuroid  skeletal  ontogeny  and  homol- 
ogies,  79 

HERMAN,  W.  S.  See  P.  D.  Pezalla,  148 

Hermissenda  crassicornis,  cultivation,  430 

HINES,  A.  H.  Reproduction  in  three  species 
of  intertidal  barnacles  from  central  Cali- 
fornia, 262 

Histology  of  holothurian  gonadial  tissues,  68 

Holothuroidea,  gonadal  development  in,  68 

Homarus,  juvenile  muscle  development,  55 
mouthparts  and  setae  of  larvae,  383 

Horseshoe  crab,  central  nervous  system  pep- 
tides,  148 

Host  specificity,  of  nemertean  parasite,  121 

HOVE,  H.  A.  TEN  AND  J.  C.  A.  WEERDENBUKG. 
A  generic  revision  of  the  brackish-water 
serpulid  Ficopomatus  Southern  1921  (Poly- 
chaeta :  Serpulinae),  including  Mercierella 
Fauvel  1923,  Sphaeropomatus  Tread  well 
1934,  Mercierellopsis  Rioja  1945  and  Neo- 
pomatus  Pillai  1960,  96 

HUMPHRIES,  A.  A.,  JR.  See  E.  M.  del  Pino, 
198 

Hydr actinia  echinata,  in  association  with  Pa- 
gurus  acadianus  and  as  prey  for  Cuthona 
nana,  157 

Hydrocorals,  prostaglandin  synthetase,  440 

Hydroid,  life  cycle,  485 

Hydromedusae,  of  Corymorphidae,  485 

Hyperglycemia,  in  Orconectes,  caused  by  CNS 
extracts  from  Limulus,  148 


Innervation,  of  crustacean  auxiliary  heart,  497 
Intertidal    development,    Nassarius    obsoletus, 

282 
Invertebrate  hearts,  anatomy,  4()7 

reproduction,  Nassarius  obsoletus,  282 
Ionic  regulation,  in  Callianassidae,  409 


K 


KAWAMURA,  K.  See  M.  Nakauchi,  453 
KAYNE,  M.  See  D.  E.  Morse,  440 
KNIGHT,  A.  \V.  See  D.  A.  Armstrong,  15 


KURIS,  A.  M.  Life  cycle,  distribution  and 
abundance  of  Carcinonemertes  epialti,  a  ne- 
mertean egg  predator  of  the  shore  crab 
Hemigrapsus  oregonensis,  in  relation  to 
host  size,  reproduction,  and  molt  cycle, 
121 


Laboratory  culture,  of  starfish  larvae,  32 
LANG,  F.  See  C.  K.  Govind,  55 
Laomediidae,  larval  development,  241 
Larvae,  Homarus  americanus,  mouthparts  and 

setae,  383 
Larval  development,  of  Naushonia,  241 

of    Stichaster    australis    and     Coscinasteria: 

calamaria,  32 

Leptosynapta  tennis,  reproductive  cycle  of,  68 
Life  cycle,  of  Corymorpha  bigelowi,  485 

of  Hermissenda  crassicornis,  430 

of  Panagrellus  silusiae,  374 

Limulus,  central  nervous  system  peptides,  148 
Linens  ruber,  nutrition,  213 
Littorina  irrorata,  respiration,  322 
Lobster  chelipeds,  development  in  juveniles,  55 

mouthparts  and  setae  of  larvae,  383 

muscle,  in  juveniles,  55 
Locomotor  activity,  in  frog,  302 
Lunar  periodicity,  in  frog  locomotor  activity, 

302 

Lymphomyeloid  tissue,  in  rectal  gland  of 
freshwater  stingray,  508 

M 

Macrobrachium  rosenbergii,  effect  of  ionized 
and  un-ionized  ammonia  on  survival  and 
growth,  15 

Marsupial  frogs,  multinucleate  oogenesis,   198 

Media,  particulate,  for  culture  of  filter- 
feeders,  47 

Mercierella,  synonymized  with  Ficopomatus, 
96,  154 

Mercierellopsis,  synonymized  with  Ficopoma- 
tus, 96 

Metamorphosis,  in  starfish,  32 
of  Hermissenda  crassicornis,  430 
of  the  nudibranch  Cuthona  nana,  157 

MICKEL,  T.  J.  AND  J.  J.  CHILDRESS.  The 
effect  of  pH  on  oxygen  consumption  and 
activity  in  the  bathypelagic  mysid  Gna- 
thopliausia  ingens,  138 

Minna,  particulate  media  for  culture  of,  47 

Morphology,  of  the  setae  of  larval  lobsters,  383 

MORSE,  D.  E.,  M.  KAYNE,  M.  TIDYMAN,  AND 
S.  ANDERSON.  Capacity  for  biosynthesis 
of  prostaglandin-related  compounds:  dis- 
tribution and  properties  of  the  rate- 
limiting  enzyme  in  hydrocorals,  gorgo- 


520 


INDEX 


nians,    and    other    coelenteratcs    of    the 

Caribbean  and  Pacific,  440 
Mouthparls,  of  larval  lobsters,  383 
Mud  shrimp,  larval  development,  241 
osmotic  and  ionic  regulation  in,  409 
Multinucleate    oogenesis,    in    marsupial    frogs, 

198 

Muscle  development,  in  juvenile  lobsters,  55 
Myeloid   lobes,   in   rectal   gland   of  freshwater 

stingray,  508 
Mysid,  bathypclagic,  effect  of  pH   on  oxygen 

consumption,  138 

N 

NAKAUCHI,  M.  AND  K.  KAWAMURA.  Additional 

experiments  on   the   behavior   of  buds   in 

the  ascidian,  Aplidium  multiplicatum,  453 

Nassarius  obsoletus  and  N.  trivittatus,  rates  of 

water  loss  from  egg  capsules  of,  282 
Naushonia,  larval  development  241 
Nematode,  acid  phosphatase  in  life  cycle,  374 
Nemertean    biology,    in    relation    to    parasite 

ecology,  121 

Neopomatus,  synonymized  with  Ficopomatus,  96 
Neuroendocrinological  studies,  of  Limuhis,  148 
Nudibranch,  development  and  association  with 

a  hydroid  and  hermit  crab,  157 
Nutrition,  in  Linens  rtiber,  213 


OAKS,  J.  A.  See  F.  M.  Fisher,  Jr.,  213 
Odontocete  cetacean   group  organization,   348 
Oogenesis,   multinucleate,   in   marsupial   frogs, 

198 

Ophiuroid,  skeletal  ontogeny  and  phylogeny,  79 
Opisthobranch  mollusc,  cultivation,  430 
Oreaster  reticulatus,  functional  morphology  of 

digestive  system,  1 
Osmoregulation,  in  Callianassidae,  409 

of  elasmobranchs,  508 
Oxygen    consumption,    effect    of    pH    on,    in 

Gnathophausia  ingens,  138 
in  crab,  188 


Pagurus  acadianus,  associated  with  Hydractinia 

and  Cuthona  nana,  157 
PALINCSAR,  E.  E.  See  G.  N.  Doering,  374 
PanagreUus   silusiae,   acid   phosphatase   in    life 

cycle  of,  374 

Panopeus  herbstii,  respiration,  188 
Panulirus  argus,  taurine  receptors,  226 
Parasite  ecology,  of  nemertean,  121 
PECHENIK,    J.    A.    Adaptations    to    intertidal 
development :   studies  on   Nassarius  obso- 
letus, 282 


Peptides,     from     central     nervous    system     of 

Limulus,  148 

PERRON,   F.   E.   Seasonal   burrowing  behavior 
and     ecology     of    Aporrhais     occidentalis 
(Gastropoda:  Strombacea),  463 
PEZALLA,   P.   D.,   R.   M.   DORES,  AND  W.  S. 
HERMAN.  Separation  and  partial  purifica- 
tion  of   central   nervous  system   peptides 
from    Limulus    polyphemus    with    hyper- 
glycemic     and     chromatophorotropic     ac- 
tivity in  crustaceans,  148 
pH,     effects    on     oxygen     consumption     in     a 

bathypelagic  mysid,  138 
Phoronids,  food  resources,  feeding  and  spatial 

pattern  in,  472 
Plwronopsis    viridis,     food-resources,     feeding, 

and  spatial  pattern  in,  472 
Physiological  flexibility,  in  copepod,  177 
Porpoise,  group  organization,  348 
Potamotrygon,  rectal  gland,  508,  154 
Predator-prey    association,    of    a    nudibranch 

and  a  hydroid,  157,  154 
Prostaglandin  synthetase,  in  corals,  440 
PROVASOLI,  L.  See  D.  E.  Conklin,  47 
PROVENZANO,  A.  J.,  JR.  See  J.  W.  Goy,  241 
PRUSCH,    R.    D.    AND    C.    HALL.    Diffusional 
water    permeability    in     selected    marine 
bivalves,  292 
Pyloric  stomach,  in  Oreaster  reticulatus,  1 

R 

Rana  pipiens,  light-dark  cycle,  lunar  perio- 
dicity, 302 

Receptor  specificity,  in  spiny  lobster,  226 

Rectal  gland,  of  freshwater  stingrays,  508 

Rectomyeloid  bodies,  of  rectal  gland  in  fresh- 
water stingrays,  508 

REES,  J.  T.  See  C.  Sassaman,  485 

Reproduction,  in  barnacles,  262 
in  Holothuroidea,  68 
in  marsupial  frogs,  198 
of  Nassarius  obsoletus,  282 

Respiration,  crab,  188 
in  Littorina  irrorata,  322 
in  mysid,  effects  of  pH,  138 

Rheotaxis,  in  Biomphalaria  glabrata,  361 

Rhynchocoelan  nutrition,  213 

RIVEST,  B.  R.  Development  of  the  eolid 
nudibranch  Cuthona  nana  (Alder  and 
Hancock,  1842),  and  its  relationship  with 
a  hydroid  and  hermit  crab,  157 

ROBERTSON,  D.  R.  The  light-dark  cycle  and 
a  nonlinear  analysis  of  lunar  perturba- 
tions and  barometric  pressure  associated 
with  the  annual  locomotor  activity  of  the 
frog,  Rana  pipiens,  302 

RONAN,  T.  E.,  JR.  Food-resources  and  the 
influence  of  spatial  pattern  on  feeding  in 
the  phoronid  Phoronopsis  viridis,  472 


INDEX 


521 


Salinity  tolerance,  of  callianassid  mud  shrimps, 

409 

Sand  dollars,  uptake  of  amino  acids,  335 
Sarcomeres,   short  and   long,   in   lobster  cheli- 

peds,  55 
SASSAMAN,  C.  AND  J.  T.  REES.  The  life  cycle 

of     Corymorpha     (=  Euphysora)     bigelowi 

(Maas,    1905)   and   its  significance   in   the 

systematics     of     corymorphid     hydrome- 

dusae,  485 
Seasonal  acclimatization,  in  Littorina  irrorata, 

322 

Sea-star,  Oreaster  reticulatus,  digestive  system,  1 
Serpulidae,      distribution      of      brackish-water 

species,  96 
Setae,   on    the   mouthparts   of   larval   lobsters, 

383 
Sexual    differences,    in    temperature    tolerance, 

in  a  copepod,  177 
SHIRLEY,  T.  C.,  G.  J.   DENOUX,  AND  W.   B. 

STICKLE.  Seasonal  respiration  in  the  marsh 

periwinkle,  Littorina  irrorata,  322 
Shrimp  larvae,  effect  of  ionized  and  un-ionized 

ammonia  on  survival  and  growth,  15 
Skeletal  development,  in  Amphioplus  abditus, 

79 
Snail,    freshwater,    chemoreception    and    rheo- 

taxis  in,  361 

metabolism,  Littorina  irrorata,  322 
Spatial  pattern,  influence  on  feeding  in  Phoro- 

nopsis  viridis,  472 
Specificity     of     taurine     receptors,     in     spiny 

lobster,  226 
Sphaeropomatus,  synonymized  with  Ficopoma- 

tus,  96 

Spiny  lobster,  taurine  receptors,  226 
Starfish  larval  development,  32 
STEINACKER,  A.  The  anatomy  of  the  decapod 

crustacean  auxiliary  heart,  497 
STEPHENS,  G.  C.,  M.  J.  VOLK,  S.  H.  WRIGHT, 

AND    P.    S.     BACKLUND.    Transepidermal 

accumulation  of  naturally  occurring  amino 

acids   in   the   sand   dollar,   Dendraster  e.\- 

centriais,  335 
Stichaster  austral  is,  (It-script  ions  of  larvae,  32 


STICKLE,  W.  B.  See  T.  C.  Shirley,  322 
Stingrays,  rectal  gland,  508 
Surfacing  associations,  porpoise,  348 
Synthetase,  prostaglandin,  in  corals,  440 
Systematics,    of    corymorphid    hydromedusae, 
485 


Taurine  receptors,  in  spiny  lobster,  226 
Temperature  tolerance,  in  copepod,  177 
Tetradita  squamosa,  reproduction,  262 
Thalassinidea,  larval  development,  241 
THORSON,  T.  B.,  R.  M.  WOTTON,  AND  T.  A. 
GEORGI.  Rectal  gland  of  freshwater  sting- 
rays, Potamotrygon  spp.   (Chondrichthyes : 
Potarnotrygonidae),  508 
TIDYMAN,  M.  See  D.  E.  Morse,  440 
Tiedemann's  pouch,  in  Oreaster  reticulatus,  1 
Toxicity,  ammonia,  to  larval  shrimp,  15 
Transepidermal    transport   of  amino   acids,    in 

Dendraster  excentricus,  335 
Transport  sites,   for  organic  nutrients,   in  the 

epidermis  of  a  rhynchocoelan,  213 
Tursiops  truncatus,  group  organization,  348 

U 

Uca  pugilator,  respiration,  188 
Ultrastructure,  of  Linens  ruber,  213 


Veliger,  of  Hermissenda  crassicornis,  430 
VERNBERG,  F.  J.  See  R.  F.  Dame,  188 
Vertical  zonation,  bivalve,  292 
YOLK,  M.  J.  See  G.  C.  Stephens,  335 

W 

WRIGHT,  S.  H.  See  G.  C.  Stephens,  335 
WOTTON,  R.  M.  See  T.  B.  Thorson,  508 
WEERDENBURG,  J.  C.  A.  See  H.  A.  ten  Hove,  96 
Water  permeability,  bivalve,  292 
WURSIG,    B.   Occurrence  and   group   organiza- 
tion    of     Atlantic     bottlenose     porpoises 
(Tursiops  truncatus)  in  an  Argentine  Bay, 
348 


Continued  from  Cover  Two 

of  biological  journal  titles  is  that  published  each  year  by  BIOLOGICAL  ABSTRACTS  (BIOSIS  List  of 
Serials;  most  recent  issue,  1976).  Foreign  authors,  and  others  who  are  accustomed  to  use 
THE  WORLD  LIST  OF  SCIENTIFIC  PERIODICALS,  may  find  a  booklet  published  by  the  Biological 
Council  of  the  U.K.  (obtainable  from  the  Institute  of  Biology,  41  Queen's  Gate,  London,  S.W.7, 
England,  U.K.  at  £0.65  or  $1.75)  useful,  since  it  sets  out  the  WORLD  LIST  abbreviations  for  most 
biological  journals  with  notes  of  the  USASI  abbreviations  where  these  differ.  CHEMICAL  AB- 
STRACTS publishes  quarterly  supplements  of  additional  abbreviations.  The  following  points  of 
reference  style  for  THE  BIOLOGICAL  BULLETIN  differ  from  USASI  (or  modified  WORLD  LIST) 
usage: 

A.  Journal  abbreviations,  and  book  titles,  all  underlined  (for  italics) 

B.  All  components  of  abbreviations  with  initial  capitals  (not  as  European  usage  in  WORLD 
LIST  e.g.  J.  Cell.  Comp.  Physiol.  NOT  J.  cell.  comp.  Physiol.) 

C.  All  abbreviated  components  must  be  followed  by  a  period,  whole  word  components 
must  not  (not  strictly  as  USASI  usage,  i.e.  J.  Cancer  Res.) 

D.  Space  between  all  components  (e.g.  J.  Cell.  Comp.  Physiol.  not  J.Cell.Comp. Physiol.) 

E.  We  strongly  recommend  that  more  unusual  words  in  journal  titles  be  spelled  out  in  full, 
rather  than  employing  lengthy,  peculiar  "abbreviations"  or  new  abbreviations  invented  by  the 
author.     For  example,  use  Rit  Visindafjelags  Islendinga  without  abbreviation.     Even  in  more 
familiar  languages,  Z.   Vererbungslehre  is  preferred  to  Z.  VerebLehre  (WORLD  LIST)  or  Z.  Verer- 
bungsl.   (USASI).     Accurate  and  complete  communication  of  the  reference  is  more  important  than 
minor  savings  in  printing  costs. 

F.  All  single  word  journal  titles  in  full  (e.g.  Veliger,  Ecology,  Brain). 

G.  The  order  of  abbreviated  components  should  be  the  same  as  the  word  order  of  the  com- 
plete title  (i.e.  Proc.  and  Trans,  placed  where  they  appear,  not  transposed  as  in  some  BIOLOGICAL 
ABSTRACTS  listings). 

H.  Spell  out  London,  Tokyo,  Paris,  Edinburgh,  Lisbon,  etc.  where  part  of  journal  title. 
I.  Series  letters  etc.  immediately  before  volume  number. 

J.  A  few  well-known  international  journals  in  their  preferred  forms  rather  than  WORLD 
LIST  or  USASI  usage  (e.g.  Nature,  Science,  Evolution  NOT  Nature,  Land.;  Science,  N.  Y.;  Evolution, 
Lancaster,  Pa.) 

K.  The  correct  abbreviation  for  THE  BIOLOGICAL  BULLETIN  is  Biol.  Bull. 

5.  Figures.     The  dimensions  of  the  printed  page,  5  by  7|  inches,  should  be  kept  in  mind  in 
preparing  figures  for  publication.     Illustrations  should  be  large  enough  so  that  all  details  will  be 
clear  after  appropriate  reduction.     Explanatory  matter  should  be  included  separately  in  legends 
as  far  as  possible,  although  the  axes  should  always  be  numbered  and  identified  on  the  illustration 
itself.     Figures  should  be  prepared  for  reproduction  as  either  line-cuts  or  halftones;  no  other 
methods  will  be  used.     Figures  to  be  reproduced  as  line-cuts  should  be  drawn  in  black  ink  on  white 
paper,  good  quality  tracing  cloth  or  plastic,  or  blue-lined  coordinate  paper;  those  to  be  reproduced 
as  halftones  should  be  mounted  on  board,  and  both  designating  numbers  or  letters  and  scale-bars 
should  be  affixed  directly  on  the  figures.     We  recommend  that  halftones  submitted  to  us  be 
mounted  prints  made  at  about  1£  times  the  linear  dimensions  of  the  final  printing  desired  (the 
actual  best  reductions  are  achieved  from  copy  in  the  range  from  H  to  2  times  the  linear  dimen- 
sions).    As  regards  line-blocks,  originals  can  be  designed  for  even  greater  reductions  but  are  best 
in  the  range  lj  to  3  times.     All  figures  should  be  numbered  in  consecutive  order,  with  no  distinc- 
tion between  text  and  plate-figures.     The  author's  name  should  appear  on  the  reverse  side  of  all 
figures,  and  the  inked  originals  for  line-blocks  must  be  submitted  for  block-making. 

6.  Mailing.     Manuscripts  should  be  packed  flat.     All  illustrations  larger  than  8£  by  1 1  inches 
must  be  accompanied  by  photographic  reproductions  or  tracings  that  may  be  folded  to  page  size. 

Reprints.  Reprints  may  be  obtained  at  cost;  approximate  prices  will  be  furnished  by  the 
Managing  Editor  upon  request. 


CONTENTS 


BOUSFIELD,  J.  D. 

Rheotaxis  and  chemoreception  in  the  freshwater  snail  Biomphalaria 
glabrata  (Say) :  estimation  of  the  molecular  weights  of  active  factors  361 

DOERING,  G.  N.  AND  E.  E.  PALINCSAR 

Acid  phosphatase  during  the  life  cycle  of  the  nematode,  Panagrellus 
silusiae 374 

FACTOR, JAN  ROBERT 

Morphology  of  the  mouthparts  of  larval  lobsters,  Homarus  ameri- 
canus  (Decapoda :  Nephropidae),  with  special  emphasis  on  their 
setae 383 

FELDER,  DARRYL  L. 

Osmotic  and  ionic  regulation  in  several  western  Atlantic  Calli- 
anassidae  (Crustacea,  Decapoda,  Thalassinidea) 409 

HARRIGAN,  JUNE  F.  AND  DANIEL  L.  ALKON 

Larval  rearing,  metamorphosis,  growth  and  reproduction  of  the 
eolid  nudibranch,  Hermissenda  crassicornis  (Eschscholtz,  1831) 
(Gastropoda :  Opisthobranchia) 430 

MORSE,  DANIEL  E.,   MARK  KAYNE,   MARK  TIDYMAN,  AND   SHANE 
ANDERSON 

Capacity  for  biosynthesis  of  prostaglandin-related  compounds : 
distribution  and  properties  of  the  rate-limiting  enzyme  in  hydro- 
corals,  gorgonians,  and  other  coelenterates  of  the  Caribbean  and 
Pacific 440 

NAKAUCHI,  MITSUAKI  AND  KAZUO  KAWAMURA 

Additional  experiments  on  the  behavior  of  buds  in  the  ascidian, 
Aplidium  multiplication 453 

PERRON, FRANK  E. 

Seasonal  burrowing  behavior  and  ecology  of  Aporrhais  occidentalis 
(Gastropoda :  Strombacea) 463  •'' 

RONAN,  THOMAS  E.,  JR. 

Food-resources  and  the  influence  of  spatial  pattern  on  feeding  in 

the  phoronid  Phoronopsis  viridis 472 

SASSAMAN,  CLAY  AND  JOHN  T.  REES 

The  life  cycle  of  Corymorpha  (  =  Euphysora)  bigelowi  (Maas,  1905) 
and  its  significance  in  the  systematics  of  corymorphid  hydromedusae -  485 

STEINACKER,  A. 

The  anatomy  of  the  decapod  crustacean  auxiliary  heart 497 

THORSON,  THOMAS  B.,  ROBERT  M.  WOTTON,  AND  TODD  A.  GEORGI 

Rectal  gland  of  freshwater  stingrays,  Potamotrygon  spp.  (Chon- 
drichthyes :  Potamotrygonidae) 508 

INDEX  TO  VOLUME  154.  517 


LIBRARY  BINDING  CO.  INC, 

JAN        1979 


f '  ~  r  r  "i  r, 


MASS. 


MBL  WHOI    LIBRARY 


IJH    1B1R    3